1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.525 2007/08/22 14:20:28 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Another patch from the Sieve maintainer.
11 PH/02 When an IPv6 address is converted to a string for single-key lookup
12 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
13 dots are used instead of colons so that keys in lsearch files need not
14 contain colons. This was done some time before quoting was made available
15 in lsearch files. However, iplsearch files do require colons in IPv6 keys
16 (notated using the quote facility) so as to distinguish them from IPv4
17 keys. This meant that lookups for IP addresses in host lists did not work
18 for iplsearch lookups.
20 This has been fixed by arranging for IPv6 addresses to be expressed with
21 colons if the lookup type is iplsearch. This is not incompatible, because
22 previously such lookups could never work.
24 The situation is now rather anomolous, since one *can* have colons in
25 ordinary lsearch keys. However, making the change in all cases is
26 incompatible and would probably break a number of configurations.
28 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
31 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
32 conversion specification without a maximum field width, thereby enabling
33 a rogue spamd server to cause a buffer overflow. While nobody in their
34 right mind would setup Exim to query an untrusted spamd server, an
35 attacker that gains access to a server running spamd could potentially
36 exploit this vulnerability to run arbitrary code as the Exim user.
38 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
39 $primary_hostname instead of what libspf2 thinks the hosts name is.
41 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
42 a directory entry by the name of the lookup key. Previously, if a
43 symlink pointed to a non-existing file or a file in a directory that
44 Exim lacked permissions to read, a lookup for a key matching that
45 symlink would fail. Now it is enough that a matching directory entry
46 exists, symlink or not. (Bugzilla 503.)
48 PH/03 The body_linecount and body_zerocount variables are now exported in the
51 PH/04 Added the $dnslist_matched variable.
53 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
54 This means they are set thereafter only if the connection becomes
57 PH/06 Added the client_condition to authenticators so that some can be skipped
58 by clients under certain conditions.
60 PH/07 The error message for a badly-placed control=no_multiline_responses left
61 "_responses" off the end of the name.
63 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
65 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
66 (without spaces) instead of just copying the configuration text.
68 PH/10 Added the /noupdate option to the ratelimit ACL condition.
70 PH/11 Added $max_received_linelength.
72 PH/12 Added +ignore_defer and +include_defer to host lists.
74 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
75 way in which PCRE > 7.0 is built.
77 PH/14 Implemented queue_only_load_latch.
79 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
80 MAIL command. The effect was to mangle the value on 64-bit systems.
82 PH/16 Another patch from the Sieve maintainer.
84 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
86 PH/18 If a system quota error occurred while trying to create the file for
87 a maildir delivery, the message "Mailbox is full" was not appended to the
88 bounce if the delivery eventually timed out. Change 4.67/27 below applied
89 only to a quota excession during the actual writing of the file.
91 PH/10 It seems that peer DN values may contain newlines (and other non-printing
92 characters?) which causes problems in log lines. The DN values are now
93 passed through string_printing() before being added to log lines.
99 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
100 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
103 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
104 issue a MAIL command.
106 PH/02 In an ACL statement such as
108 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
110 if a client was not listed at all, or was listed with a value other than
111 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
112 the condition was not true (as it should be), so access was not denied.
113 The bug was that the ! inversion was incorrectly passed on to the second
114 item. This has been fixed.
116 PH/03 Added additional dnslists conditions == and =& which are different from
117 = and & when the dns lookup returns more than one IP address.
119 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
120 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
122 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
123 FSYNC, which compiles an option called disable_fsync that allows for
124 bypassing fsync(). The documentation is heavily laced with warnings.
126 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
128 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
129 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
130 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
131 including adding "make clean"; (3) Added -fPIC when compiling the test
132 dynamically loaded module, to get rid of a warning.
134 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
135 message fails, move_frozen_messages = true and ignore_bounce_errors_after
136 = 0s. The bug is otherwise harmless.
138 PH/07 There was a bug in the dovecot authenticator such that the value of
139 $auth1 could be overwritten, and so not correctly preserved, after a
140 successful authentication. This usually meant that the value preserved by
141 the server_setid option was incorrect.
143 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
145 PH/09 Installed PCRE release 7.0.
147 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
148 run for batched SMTP input. It is now run at the start of every message
149 in the batch. While fixing this I discovered that the process information
150 (output by running exiwhat) was not always getting set for -bs and -bS
151 input. This is fixed, and it now also says "batched" for BSMTP.
153 PH/11 Added control=no_pipelining.
155 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
156 patch, slightly modified), and move the expansion of helo_data till after
157 the connection is made in the smtp transport (so it can use these
160 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
162 PH/14 Added log_selector = +pid.
164 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
166 PH/16 Add ${if forany and ${if forall.
168 PH/17 Added dsn_from option to vary the From: line in DSNs.
170 PH/18 Flush SMTP output before performing a callout, unless control =
171 no_callout_flush is set.
173 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
174 was true (the default) a successful delivery failed to delete the retry
175 item, thus causing premature timeout of the address. The bug is now
178 PH/20 Added hosts_avoid_pipelining to the smtp transport.
180 PH/21 Long custom messages for fakedefer and fakereject are now split up
181 into multiline reponses in the same way that messages for "deny" and
182 other ACL rejections are.
184 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
185 with slight modification.
187 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
188 draft, changing the syntax and factoring some duplicate code".
190 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
191 for deliveries of the second and subsequent messages over the same SMTP
194 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
195 ${reduce, with only minor "tidies".
197 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
199 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
200 expansion side effects.
202 PH/27 When a message times out after an over-quota error from an Exim-imposed
203 quota, the bounce message says "mailbox is full". This message was not
204 being given when it was a system quota that was exceeded. It now should
207 MH/03 Made $recipients available in local_scan(). local_scan() already has
208 better access to the recipient list through recipients_list[], but
209 $recipients can be useful in postmaster-provided expansion strings.
211 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
212 in the case of a MAIL command with additional options following the
213 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
214 were accidentally chopped off.
216 PH/29 SMTP synchronization checks are implemented when a command is read -
217 there is a check that no more input is waiting when there shouldn't be
218 any. However, for some commands, a delay in an ACL can mean that it is
219 some time before the response is written. In this time, more input might
220 arrive, invalidly. So now there are extra checks after an ACL has run for
221 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
222 pipelining has not been advertised.
224 PH/30 MH's patch to allow iscntrl() characters to be list separators.
226 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
227 returned in the SMTP response when smtp_return_error_details was false.
230 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
231 instead of the C I/O that was originally supplied, because problems were
234 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
235 Exim which did not show up earlier: it was assuming that a call to
236 SSL_CTX_set_info_callback() might give an error value. In fact, there is
237 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
238 was a macro that became an assignment, so it seemed to work. This has
239 changed to a proper function call with a void return, hence the compile
240 error. Exim's code has been fixed.
242 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
245 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
247 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
249 PH/37 If a message is not accepted after it has had an id assigned (e.g.
250 because it turns out to be too big or there is a timeout) there is no
251 "Completed" line in the log. When some messages of this type were
252 selected by exigrep, they were listed as "not completed". Others were
253 picked up by some special patterns. I have improved the selection
254 criteria to be more general.
256 PH/38 The host_find_failed option in the manualroute router can now be set
257 to "ignore", to completely ignore a host whose IP address cannot be
258 found. If all hosts are ignored, the behaviour is controlled by the new
259 host_all_ignored option.
261 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
262 homing or because of multiple MX records with /mx) generated more than
263 one IP address, and the following item turned out to be the local host,
264 all the secondary addresses of the first item were incorrectly removed
265 from the list, along with the local host and any following hosts (which
266 is what is supposed to happen).
268 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
269 whoever called Exim into the -H file. In the case of the daemon it was
270 behaving confusingly. When first started, it used values for whoever
271 started the daemon, but after a SIGHUP it used the Exim user (because it
272 calls itself on a restart). I have changed the code so that it now always
275 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
276 message are rejected with the same error (e.g. no authentication or bad
277 sender address), and a DATA command is nevertheless sent (as can happen
278 with PIPELINING or a stupid MUA), the error message that was given to the
279 RCPT commands is included in the rejection of the DATA command. This is
280 intended to be helpful for MUAs that show only the final error to their
283 PH/42 Another patch from the Sieve maintainer.
285 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
286 Eximstats - Fixed some broken HTML links and added missing column headers
288 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
289 columns for Rejects, Temp Rejects, Ham, and Spam rows.
291 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
293 PH/43 Yet another patch from the Sieve maintainer.
295 PH/44 I found a way to check for a TCP/IP connection going away before sending
296 the response to the final '.' that terminates a message, but only in the
297 case where the client has not sent further data following the '.'
298 (unfortunately, this is allowed). However, in many cases there won't be
299 any further data because there won't be any more messages to send. A call
300 to select() can be used: if it shows that the input is "ready", there is
301 either input waiting, or the socket has been closed. An attempt to read
302 the next input character can distinguish the two cases. Previously, Exim
303 would have sent an OK response which the client would never have see.
304 This could lead to message repetition. This fix should cure that, at
305 least in a lot of common cases.
307 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
308 advertised in response to EHLO.
314 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
315 fixed by 4.65/MH/01 (is this a record?) are fixed:
317 (i) An empty string was always treated as zero by the numeric comparison
318 operators. This behaviour has been restored.
320 (ii) It is documented that the numeric comparison operators always treat
321 their arguments as decimal numbers. This was broken in that numbers
322 starting with 0 were being interpreted as octal.
324 While fixing these problems I realized that there was another issue that
325 hadn't been noticed. Values of message_size_limit (both the global option
326 and the transport option) were treated as octal if they started with 0.
327 The documentation was vague. These values are now always treated as
328 decimal, and I will make that clear in the documentation.
334 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
335 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
338 MH/01 Don't check that the operands of numeric comparison operators are
339 integers when their expansion is in "skipping" mode (fixes bug
340 introduced by 4.64-PH/07).
342 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
343 child addresses, Exim now panics and dies. Previously, because the count
344 is held in a short int, deliveries were likely to be lost. As such a
345 large number of recipients for a single message is ridiculous
346 (performance will be very, very poor), I have chosen to impose a limit
347 rather than extend the field.
353 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
354 leftover -K file (the existence of which was triggered by #402).
355 While we were at it, introduced process PID as part of the -K
356 filename. This should rule out race conditions when creating
359 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
360 processing considerably. Previous code took too long for large mails,
361 triggering a timeout which in turn triggers #401.
363 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
364 in the DK code in transports.c. sendfile() is not really portable,
365 hence the _LINUX specificness.
367 TF/01 In the add_headers option to the mail command in an Exim filter,
368 there was a bug that Exim would claim a syntax error in any
369 header after the first one which had an odd number of characters
372 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
373 callout verification, Exim cached a "reject" for the entire domain. This
374 is correct for most verifications, but it is not correct for a recipient
375 verification with use_sender or use_postmaster set, because in that case
376 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
377 case of MAIL FROM:<> rejection from other early rejections (e.g.
378 rejection of HELO). When verifying a recipient using a non-null MAIL
379 address, the cache is ignored if it shows MAIL FROM:<> rejection.
380 Whatever the result of the callout, the value of the domain cache is
381 left unchanged (for any other kind of callout, getting as far as trying
382 RCPT means that the domain itself is ok).
384 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
387 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
388 manifest itself as EPIPE rather than ECONNECT. When tidying away a
389 session, the daemon ignores ECONNECT errors and logs others; it now
390 ignores EPIPE as well.
392 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
393 (quoted-printable decoding).
395 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
396 later the small subsequent patch to fix an introduced bug.
398 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
400 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
402 PH/08 An error is now given if message_size_limit is specified negative.
404 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
405 to be given (somewhat) arbitrary names.
407 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
410 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
411 miscellaneous code fixes
413 PH/10 Added the log_reject_target ACL modifier to specify where to log
416 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
417 hostname. This is wrong, because it relates to the incoming message (and
418 probably the interface on which it is arriving) and not to the outgoing
419 callout (which could be using a different interface). This has been
420 changed to use the value of the helo_data option from the smtp transport
421 instead - this is what is used when a message is actually being sent. If
422 there is no remote transport (possible with a router that sets up host
423 addresses), $smtp_active_hostname is used.
425 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
426 tweaks were necessary in order to get it to work (see also 21 below):
427 (a) The code assumed that strncpy() returns a negative number on buffer
428 overflow, which isn't the case. Replaced with Exim's string_format()
430 (b) There were several signed/unsigned issues. I just did the minimum
431 hacking in of casts. There is scope for a larger refactoring.
432 (c) The code used strcasecmp() which is not a standard C function.
433 Replaced with Exim's strcmpic() function.
434 (d) The code set only $1; it now sets $auth1 as well.
435 (e) A simple test gave the error "authentication client didn't specify
436 service in request". It would seem that Dovecot has changed its
437 interface. Fortunately there's a specification; I followed it and
438 changed what the client sends and it appears to be working now.
440 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
443 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
444 address A is aliased to B and C, where B exists and C does not. Without
445 -v the output is "A verified" because verification stops after a
446 successful redirection if more than one address is generated. However,
447 with -v the child addresses are also verified. Exim was outputting "A
448 failed to verify" and then showing the successful verification for C,
449 with its parentage. It now outputs "B failed to verify", showing B's
450 parentage before showing the successful verification of C.
452 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
453 look up a TXT record in a specific list after matching in a combined
456 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
457 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
458 they consult the DNS. I had assumed they would set it the way they
459 wanted; and indeed my experiments on Linux seem to show that in some
460 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
461 To be on the safe side, however, I have now made the interface to
462 host_find_byname() similar to host_find_bydns(), with an argument
463 containing the DNS resolver options. The host_find_byname() function now
464 sets these options at its start, just as host_find_bydns() does. The smtp
465 transport options dns_qualify_single and dns_search_parents are passed to
466 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
467 of host_find_byname() use the default settings of RES_DEFNAMES
468 (qualify_single) but not RES_DNSRCH (search_parents).
470 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
471 spool_read_header() do less string testing, by means of a preliminary
472 switch on the second character of optional "-foo" lines. (This is
473 overdue, caused by the large number of possibilities that now exist.
474 Originally there were few.) While I was there, I also converted the
475 str(n)cmp tests so they don't re-test the leading "-" and the first
476 character, in the hope this might squeeze out yet more improvement.
478 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
479 flag allowing group syntax was set by the header_syntax check but not
480 turned off, possible causing trouble later; (2) The flag was not being
481 set at all for the header_verify test, causing "group"-style headers to
482 be rejected. I have now set it in this case, and also caused header_
483 verify to ignore an empty address taken from a group. While doing this, I
484 came across some other cases where the code for allowing group syntax
485 while scanning a header line wasn't quite right (mostly, not resetting
486 the flag correctly in the right place). These bugs could have caused
487 trouble for malformed header lines. I hope it is now all correct.
489 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
490 with the "reply" argument non-NULL. The code, however (which originally
491 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
492 but it didn't always do it. This confused somebody who was copying the
493 code for some other use. I have removed all the tests.
495 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
496 feature that was used to support insecure browsers during the U.S. crypto
497 embargo. It requires special client support, and Exim is probably the
498 only MTA that supported it -- and would never use it because real RSA is
499 always available. This code has been removed, because it had the bad
500 effect of slowing Exim down by computing (never used) parameters for the
501 RSA_EXPORT functionality.
503 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
504 authenticator to fail if there's a tab character in the incoming data
505 (there should never be unless someone is messing about, as it's supposed
506 to be base64-encoded). Also added, on Timo's advice, the "secured" option
507 if the connection is using TLS or if the remote IP is the same as the
508 local IP, and the "valid-client-cert option" if a client certificate has
511 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
512 authenticators. This can be used for authorization after authentication
513 succeeds. (In the case of plaintext, it servers for both authentication
516 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
517 if any retry times were supplied.
519 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
520 connection, where there is no client IP address to check. In this
521 situation, the verify now always succeeds.
523 PH/25 Applied John Jetmore's -Mset patch.
525 PH/26 Added -bem to be like -Mset, but loading a message from a file.
527 PH/27 In a string expansion for a processed (not raw) header when multiple
528 headers of the same name were present, leading whitespace was being
529 removed from all of them, but trailing whitespace was being removed only
530 from the last one. Now trailing whitespace is removed from each header
531 before concatenation. Completely empty headers in a concatenation (as
534 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
535 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
537 PH/29 [Removed. This was a change that I later backed out, and forgot to
538 correct the ChangeLog entry (that I had efficiently created) before
539 committing the later change.]
541 PH/30 Exim was sometimes attempting to deliver messages that had suffered
542 address errors (4xx response to RCPT) over the same connection as other
543 messages routed to the same hosts. Such deliveries are always "forced",
544 so retry times are not inspected. This resulted in far too many retries
545 for the affected addresses. The effect occurred only when there were more
546 hosts than the hosts_max_try setting in the smtp transport when it had
547 the 4xx errors. Those hosts that it had tried were not added to the list
548 of hosts for which the message was waiting, so if all were tried, there
549 was no problem. Two fixes have been applied:
551 (i) If there are any address or message errors in an SMTP delivery, none
552 of the hosts (tried or untried) are now added to the list of hosts
553 for which the message is waiting, so the message should not be a
554 candidate for sending over the same connection that was used for a
555 successful delivery of some other message. This seems entirely
556 reasonable: after all the message is NOT "waiting for some host".
557 This is so "obvious" that I'm not sure why it wasn't done
558 previously. Hope I haven't missed anything, but it can't do any
559 harm, as the worst effect is to miss an optimization.
561 (ii) If, despite (i), such a delivery is accidentally attempted, the
562 routing retry time is respected, so at least it doesn't keep
563 hammering the server.
565 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
566 in ${readsocket because some servers need this prod.
568 PH/32 Added some extra debug output when updating a wait-xxx database.
570 PH/33 The hint "could be header name not terminated by colon", which has been
571 given for certain expansion errors for a long time, was not being given
572 for the ${if def:h_colon_omitted{... case.
574 PH/34 The spec says: "With one important exception, whenever a domain list is
575 being scanned, $domain contains the subject domain." There was at least
576 one case where this was not true.
578 PH/35 The error "getsockname() failed: connection reset by peer" was being
579 written to the panic log as well as the main log, but it isn't really
580 panic-worthy as it just means the connection died rather early on. I have
581 removed the panic log writing for the ECONNRESET error when getsockname()
584 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
585 runs only) independently of the message's sender address. This meant
586 that, if the 4xx error was in fact related to the sender, a different
587 message to the same recipient with a different sender could confuse
588 things. In particualar, this can happen when sending to a greylisting
589 server, but other circumstances could also provoke similar problems.
590 I have changed the default so that the retry time for these errors is now
591 based a combination of the sender and recipient addresses. This change
592 can be overridden by setting address_retry_include_sender=false in the
595 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
596 remote server are returned as part of bounce messages. This was not
597 happening for LMTP over a pipe (the lmtp transport), but now it is the
598 same for both kinds of LMTP.
600 PH/38 Despite being documented as not happening, Exim was rewriting addresses
601 in header lines that were in fact CNAMEs. This is no longer the case.
603 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
604 and queue runs started by the daemon processed all messages. This has
605 been fixed so that -R and -S can now usefully be given with -q<time>.
607 PH/40 Import PCRE release 6.7 (fixes some bugs).
609 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
611 PH/42 Give an error if -q is specified more than once.
613 PH/43 Renamed the variables $interface_address and $interface_port as
614 $received_ip_address and $received_port, to make it clear that these
615 values apply to message reception, and not to the outgoing interface when
616 a message is delivered. (The old names remain recognized, of course.)
618 PH/44 There was no timeout on the connect() call when using a Unix domain
619 socket in the ${readsocket expansion. There now is.
621 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
622 be meaningful with "accept".
624 SC/01 Eximstats V1.43
625 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
627 SC/02 Eximstats V1.44
628 Use a glob alias rather than an array ref in the generated
629 parser. This improves both readability and performance.
631 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
632 Collect SpamAssassin and rejection statistics.
633 Don't display local sender or destination tables unless
634 there is data to show.
635 Added average volumes into the top table text output.
637 SC/04 Eximstats V1.46
638 Collect data on the number of addresses (recipients)
639 as well as the number of messages.
641 SC/05 Eximstats V1.47
642 Added 'Message too big' to the list of mail rejection
643 reasons (thanks to Marco Gaiarin).
645 SC/06 Eximstats V1.48
646 Mainlog lines which have GMT offsets and are too short to
647 have a flag are now skipped.
649 SC/07 Eximstats V1.49 (Alain Williams)
650 Added the -emptyok flag.
652 SC/08 Eximstats V1.50
653 Fixes for obtaining the IP address from reject messages.
655 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
656 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
657 whitesspace changes from 4.64-PH/27
659 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
662 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
663 are found, allow negative numbers in numeric criteria)
665 JJ/06 exipick.20061117.2, added new $message_body_missing variable
667 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
668 to match changes made in 4.64-PH/43
670 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
672 PH/47 Put in an explicit test for a DNS lookup of an address record where the
673 "domain" is actually an IP address, and force a failure. This locks out
674 those revolvers/nameservers that support "A-for-A" lookups, in
675 contravention of the specifications.
677 PH/48 When a host name was looked up from an IP address, and the subsequent
678 forward lookup of the name timed out, the host name was left in
679 $sender_host_name, contrary to the specification.
681 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
682 restricted to single-key lookups, Exim was not diagnosing an error if
683 * or *@ was used with a query-style lookup.
685 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
687 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
688 long ago, but noone interested enough thought of it. Let's just say that
689 the "1.1" means that there are some new functions that weren't there at
690 some point in the past.
692 PH/51 Error processing for expansion failure of helo_data from an smtp
693 transport during callout processing was broken.
695 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
696 tested/used via the -bh/-bhc/-bs options.
698 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
699 bug, fixed in subsequent PCRE releases).
701 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
702 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
704 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
710 SC/01 Use a glob alias rather than an array ref in eximstats generated
711 parser. This improves both readability and performance.
713 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
714 Don't display local sender or destination tables in eximstats unless
715 there is data to show.
716 Added average volumes into the eximstats top table text output.
718 SC/03 Collect data on the number of addresses (recipients) as well
719 as the number of messages in eximstats.
721 TF/01 Correct an error in the documentation for the redirect router. Exim
722 does (usually) call initgroups() when daemonizing.
724 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
725 with consistent privilege compared to when running as a daemon.
727 TF/03 Note in the spec that $authenticated_id is not set for local
728 submissions from trusted users.
730 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
731 Thanks to Dean Brooks <dean@iglou.com> for the patch.
733 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
734 by adding some example configuration directives to the default
735 configuration file. A little bit of work is required to uncomment the
736 directives and define how usernames and passwords are checked, but
737 there is now a framework to start from.
739 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
740 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
741 without this. I don't know how relevant this is to other LDAP libraries.
743 PH/02 Add the verb name to the "unknown ACL verb" error.
745 PH/03 Magnus Holmgren's patch for filter_prepend_home.
747 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
749 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
750 directory not expanded when it should be if an expanded home directory
751 was set for the address (which is overridden by the transport).
753 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
756 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
757 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
758 because it is too late at that time, and has no effect.
760 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
761 security issue with \' (bugzilla #107). I could not use the
762 PQescapeStringConn() function, because it needs a PGconn value as one of
765 PH/08 When testing addresses using -bt, indicate those final addresses that
766 are duplicates that would not cause an additional delivery. At least one
767 person was confused, thinking that -bt output corresponded to deliveries.
768 (Suppressing duplicates isn't a good idea as you lose the information
769 about possibly different redirections that led to the duplicates.)
771 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
772 systems where poll() doesn't work, in particular OS X.
774 PH/10 Added more information to debugging output for retry time not reached.
776 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
777 operations in malware.c.
779 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
782 PH/13 If write_rejectlog was set false when logging was sent to syslog with
783 syslog_duplication set false, log lines that would normally be written
784 both the the main log and to the reject log were not written to syslog at
787 PH/14 In the default configuration, change the use of "message" in ACL warn
788 statements to "add_header".
790 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
791 not followed by a command (e.g. "seen endif").
793 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
794 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
797 PH/17 Added extra conditions to the default value of delay_warning_condition
801 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
802 { match{$h_precedence:}{(?i)bulk|list|junk} } \
803 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
806 The Auto-Submitted: and various List- headers are standardised, whereas I
807 don't think Precedence: ever was.
809 PH/18 Refactored debugging code in route_finduser() to show more information,
810 in particular, the error code if getpwnam() issues one.
812 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
813 This is apparently needed in addition to the PH/07 change above to avoid
814 any possible encoding problems.
816 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
817 but not after initializing Perl.
819 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
820 output them only if debugging. By default they are written stderr,
821 apparently, which is not desirable.
823 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
826 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
829 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
831 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
832 authenticated or an ident call has been made. Suppress the default
833 values for $authenticated_id and $authenticated_sender (but permit -oMai
834 and -oMas) when testing with -bh.
836 PH/24 Re-jigged the order of the tests in the default configuration so that the
837 tests for valid domains and recipients precede the DNS black list and CSA
838 tests, on the grounds that those ones are more expensive.
840 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
841 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
842 This bug exists in every version of Exim that I still have, right back to
845 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
846 However, an attempt to turn on case-sensitivity in a regex key by
847 including (?-i) didn't work because the subject string was already
848 lowercased, and the effects were non-intuitive. It turns out that a
849 one-line patch can be used to allow (?-i) to work as expected.
855 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
856 other effects) broke the use of negated acl sub-conditions.
858 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
861 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
862 "Deny" causes Exim to reject the incoming connection with a 554 error.
863 Unfortunately, if there is a major crisis, such as a disk failure,
864 tcp-wrappers gives "deny", whereas what one would like would be some
865 kind of temporary error. A kludge has been added to help with this.
866 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
867 554 error is used if errno is still zero or contains ENOENT (which occurs
868 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
871 PH/03 Add -lutil to the default FreeBSD LIBS setting.
873 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
874 errors. Otherwise a message that provokes a temporary error (when other
875 messages do not) can cause a whole host to time out.
877 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
878 addresses were routed directly to files or pipes from a redirect router.
879 File deliveries just didn't batch; pipe deliveries might have suffered
882 PH/06 A failure to get a lock for a hints database would erroneously always say
883 "Failed to get write lock", even when it was really a read lock.
885 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
886 of 0600. This has been changed to use the value of the lockfile_mode
887 option (which defaults to 0600).
889 PH/08 Applied small patch from the Sieve maintainer.
891 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
892 folder from quota calculations, a direct delivery into this folder messed
893 up the contents of the maildirsize file. This was because the regex was
894 used only to exclude .Trash (or whatever) when the size of the mailbox
895 was calculated. There was no check that a delivery was happening into an
896 excluded directory. This bug has been fixed by ignoring all quota
897 processing for deliveries into excluded directories.
899 PH/10 Added the maildirfolder_create_regex option to appendfile.
905 PH/01 The code for finding all the local interface addresses on a FreeBSD
906 system running IPv6 was broken. This may well have applied to all BSD
907 systems, as well as to others that have similar system calls. The broken
908 code found IPv4 interfaces correctly, but gave incorrect values for the
909 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
910 that it would not match correctly against @[] and not recognize the IPv6
913 PH/02 The ipliteral router was not recognizing addresses of the form user@
914 [ipv6:....] because it didn't know about the "ipv6:" prefix.
916 PH/03 Added disable_ipv6.
918 PH/04 Changed $reply_address to use the raw form of the headers instead of the
919 decoded form, because it is most often used to construct To: headers
920 lines in autoreplies, and the decoded form may well be syntactically
921 invalid. However, $reply_address has leading white space removed, and all
922 newlines turned into spaces so that the autoreply transport does not
925 PH/05 If group was specified without a user on a router, and no group or user
926 was specified on a transport, the group from the router was ignored.
928 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
929 for visible compile-time settings that can be used to change these
930 numbers, for those that want even more. Backwards compatibility with old
931 spool files has been maintained. However, going back to a previous Exim
932 release will lost any variables that are in spool files.
934 PH/07 Two small changes when running in the test harness: increase delay when
935 passing a TCP/IP connection to a new process, in case the original
936 process has to generate a bounce, and remove special handling of
937 127.0.0.2 (sic), which is no longer necessary.
939 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
940 be the same on different OS.
942 PH/09 Moved a debug statement in filter processing to avoid a race problem when
945 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
946 whether --show-vars was specified or not
948 JJ/02 exipick: Added support for new ACL variable spool format introduced
951 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
952 syntactically invalid From: or Reply-to: line, and a filter used this to
953 generate an autoreply, and therefore failed to obtain an address for the
954 autoreply, Exim could try to deliver to a non-existent relative file
955 name, causing unrelated and misleading errors. What now happens is that
956 it logs this as a hard delivery error, but does not attempt to create a
959 PH/11 The exinext utility has a -C option for testing purposes, but although
960 the given file was scanned by exinext itself; it wasn't being passed on
961 when Exim was called.
963 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
964 an end-of-file indication when reading a command response.
966 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
967 compiled. In many other places in Exim, IPv6 addresses are always
968 recognized, so I have changed this. It also means that IPv4 domain
969 literals of the form [IPV4:n.n.n.n] are now always recognized.
971 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
972 used if the router is not running as root, for example, when verifying at
973 ACL time, or when using -bh. The debugging output from this situation was
974 non-existent - all you got was a failure to exec. I have made two
977 (a) Failures to set uid/gid, the current directory, or a process leader
978 in a subprocess such as that created by queryprogram now generate
979 suitable debugging ouput when -d is set.
981 (b) The queryprogram router detects when it is not running as root,
982 outputs suitable debugging information if -d is set, and then runs
983 the subprocess without attempting to change uid/gid.
985 PH/15 Minor change to Makefile for building test_host (undocumented testing
988 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
989 additional section of a DNS packet that returns MX or SRV records.
990 Instead, it always explicitly searches for A/AAAA records. This avoids
991 major problems that occur when a DNS server includes only records of one
992 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
993 fixed another bug: if SRV records were looked up and the corresponding
994 address records were *not* found in the additional section, the port
995 values from the SRV records were lost.
997 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
998 using the correct key (the original address) when searching the retry
999 rules in order to find which one to use for generating the retry hint.
1001 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1002 adding the default one. Similarly, if it contains a Reply-To: header, the
1003 errors_reply_to option, if set, is not used.
1005 PH/19 When calculating a retry time, Exim used to measure the "time since
1006 failure" by looking at the "first failed" field in the retry record. Now
1007 it does not use this if it is later than than the arrival time of the
1008 message. Instead it uses the arrival time. This makes for better
1009 behaviour in cases where some deliveries succeed, thus re-setting the
1010 "first failed" field. An example is a quota failure for a huge message
1011 when small messages continue to be delivered. Without this change, the
1012 "time since failure" will always be short, possible causing more frequent
1013 delivery attempts for the huge message than are intended.
1014 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1016 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1017 $1, $2, $3) because the numerical variables can be reset during some
1018 expansion items (e.g. "match"), thereby losing the authentication data.
1020 PH/21 Make -bV show the size of off_t variables so that the test suite can
1021 decide whether to run tests for quotas > 2G.
1023 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1024 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1025 filecount value is greater than 2G or if a quota value is greater than 2G
1026 on a system where the size of off_t is not greater than 4, a panic error
1029 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1030 never match. The debug and -bh output now contains an explicit error
1031 message indicating a malformed IPv4 address or mask.
1033 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1034 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1035 PH/23 above applies.
1037 PH/25 Do not write to syslog when running in the test harness. The only
1038 occasion when this arises is a failure to open the main or panic logs
1039 (for which there is an explicit test).
1041 PH/26 Added the /no_tell option to "control=freeze".
1043 PH/27 If a host name lookup failed very early in a connection, for example, if
1044 the IP address matched host_lookup and the reverse lookup yielded a name
1045 that did not have a forward lookup, an error message of the form "no IP
1046 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1047 could be logged. Now it outputs the IP address instead of "NULL".
1049 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1050 allows the sender and the authenticated sender to be set when
1051 submitting a message from within Exim. Since child_open_exim() is
1052 documented for local_scan(), the new function should be too.
1054 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1055 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1056 results in an empty string is now treated as unset.
1058 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1060 PH/31 Added sender_verify_fail logging option.
1062 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1063 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1064 tidied the source and removed it altogether.
1066 PH/33 When a queue run was abandoned because the load average was too high, a
1067 log line was always written; now it is written only if the queue_run log
1068 selector is set. In addition, the log line for abandonment now contains
1069 information about the queue run such as the pid. This is always present
1070 in "start" and "stop" lines but was omitted from the "abandon" line.
1072 PH/34 Omit spaces between a header name and the colon in the error message that
1073 is given when verify = headers_syntax fails (if there are lots of them,
1074 the message gets confusing).
1076 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1077 names, as there are now some PTR records that contain slashes. This check
1078 is only to protect against broken name servers that fall over on strange
1079 characters, so the fact that it applies to all lookups doesn't matter.
1081 PH/36 Now that the new test suite is complete, we can remove some of the
1082 special code in Exim that was needed for the old test suite. For example,
1083 sorting DNS records because real resolvers return them in an arbitrary
1084 order. The new test suite's fake resolver always returns records in the
1087 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1088 bounces) except when queue_only is set, to avoid logging races between
1089 the different processes.
1091 PH/38 Panic-die if .include specifies a non-absolute path.
1093 PH/39 A tweak to the "H" retry rule from its user.
1095 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1096 a label. They prevented compilation on older perls.
1098 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1099 a warning to be raised on newish perls.
1101 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1102 on queue. Changes to match documented behaviour of showing count of
1103 messages matching specified criteria.
1105 PH/40 Changed the default ident timeout from 30s to 5s.
1107 PH/41 Added support for the use of login_cap features, on those BSD systems
1108 that have them, for controlling the resources used by pipe deliveries.
1110 PH/42 The content-scanning code uses fopen() to create files in which to put
1111 message data. Previously it was not paying any attention to the mode of
1112 the files. Exim runs with umask(0) because the rest of the code creates
1113 files with open(), and sets the required mode explicitly. Thus, these
1114 files were ending up world-writeable. This was not a big issue, because,
1115 being within the spool directory, they were not world-accessible. I have
1116 created a function called modefopen, which takes an additional mode
1117 argument. It sets umask(777), creates the file, chmods it to the required
1118 mode, then resets the umask. All the relevant calls to fopen() in the
1119 content scanning code have been changed to use this function.
1121 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1122 to 24 hours. This avoids potential overflow problems when processing G
1123 and H retry rules. I suspect nobody ever tinkers with this value.
1125 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1127 PH/45 When the plaintext authenticator is running as a client, the server's
1128 challenges are checked to ensure they are valid base64 strings. By
1129 default, the authentication attempt is cancelled if an invalid string is
1130 received. Setting client_ignore_invalid_base64 true ignores these errors.
1131 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1132 they are received. Thus, the responses can be made to depend on the
1133 challenges. If an invalid string is ignored, an empty string is placed in
1136 PH/46 Messages that are created by the autoreply transport now contains a
1137 References: header, in accordance with RFCs 2822 and 3834.
1139 PH/47 Added authenticated_sender_force to the smtp transport.
1141 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1143 PH/49 Installed latest patch from the Sieve maintainer.
1145 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1146 was also set, the appendfile transport was unnecessarily scanning a
1147 directory of message files (e.g. for maildir delivery) to find the count
1148 of files (along with the size), even though it did not need this
1149 information. It now does the scan only if it needs to find either the
1150 size of the count of files.
1152 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1154 PH/52 Two bugs concerned with error handling when the smtp transport is
1157 (i) Exim was not creating retry information for temporary errors given
1158 for individual recipients after the DATA command when the smtp transport
1159 was used in LMTP mode. This meant that they could be retried too
1160 frequently, and not timed out correctly.
1162 (ii) Exim was setting the flag that allows error details to be returned
1163 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1164 recipients that were returned after the DATA command.
1166 PH/53 This is related to PH/52, but is more general: for any failing address,
1167 when detailed error information was permitted to be returned to the
1168 sender, but the error was temporary, then after the final timeout, only
1169 "retry timeout exceeded" was returned. Now it returns the full error as
1170 well as "retry timeout exceeded".
1172 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1173 do this, and (what is worse) MTAs that accept it.
1175 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1176 will now be deprecated.
1178 PH/56 New os.c-cygwin from the Cygwin maintainer.
1180 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1181 formats (previously only available in exim formats via -bpr, -bpru,
1182 and -bpra. Now also available in native and exiqgrep formats)
1184 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1185 with very large, slow to parse queues
1187 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1189 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1191 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1192 responses to authentication challenges, though it was showing the
1193 challenges; (ii) I've removed the CR characters from the debug output for
1196 PH/58 Allow for the insertion of a newline as well as a space when a string
1197 is turned into more than one encoded-word during RFC 2047 encoding. The
1198 Sieve code now uses this.
1200 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1201 data_4xx, lost_connection, tls_required.
1203 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1204 message was being sent as an SMTP response.
1206 PH/61 Add -l and -k options to exicyclog.
1208 PH/62 When verifying, if an address was redirected to one new address, so that
1209 verification continued, and the new address failed or deferred after
1210 having set something in $address_data, the value of $address_data was not
1211 passed back to the ACL. This was different to the case when no
1212 redirection occurred. The value is now passed back in both cases.
1214 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1215 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1216 use login_cap.h, so on its own it isn't the distinguishing feature. The
1217 new name refers directly to the setclassresources() function.
1219 PH/65 Added configuration files for NetBSD3.
1221 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1223 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1224 is preferred over IPv4.
1226 PH/68 The bounce_return_message and bounce_return_body options were not being
1227 honoured for bounces generated during the reception of non-SMTP messages.
1228 In particular, this applied to messages rejected by the ACL. This bug has
1229 been fixed. However, if bounce_return_message is true and bounce_return_
1230 body is false, the headers that are returned for a non-SMTP message
1231 include only those that have been read before the error was detected.
1232 (In the case of an ACL rejection, they have all been read.)
1234 PH/69 The HTML version of the specification is now built in a directory called
1235 spec_html instead of spec.html, because the latter looks like a path with
1236 a MIME-type, and this confuses some software.
1238 PH/70 Catch two compiler warnings in sieve.c.
1240 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1241 function verify_get_ident() calls ip_connect() to connect a socket, but
1242 if the "connect()" function timed out, ip_connect() used to close the
1243 socket. However, verify_get_ident() also closes the socket later, and in
1244 between Exim writes to the log, which may get opened at this point. When
1245 the socket was closed in ip_connect(), the log could get the same file
1246 descriptor number as the socket. This naturally causes chaos. The fix is
1247 not to close the socket in ip_connect(); the socket should be closed by
1248 the function that creates it. There was only one place in the code where
1249 this was missing, in the iplookup router, which I don't think anybody now
1250 uses, but I've fixed it anyway.
1252 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1253 well as to direct DNS lookups. Otherwise the handling of names in host
1254 lists is inconsistent and therefore confusing.
1260 PH/01 Two changes to the default runtime configuration:
1262 (1) Move the checks for relay_from_hosts and authenticated clients from
1263 after to before the (commented out) DNS black list checks.
1265 (2) Add control=submission to the relay_from_hosts and authenticated
1266 clients checks, on the grounds that messages accepted by these
1267 statements are most likely to be submissions.
1269 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1271 (1) Generate an error if the third argument for the ${prvs expansion is
1274 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1277 (3) Reset the variables that are obtained from the first argument of
1278 ${prvscheck and used in the second argument before leaving the code,
1279 because their memory is reclaimed, so using them afterwards may do
1282 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1283 one (it's much easier than Tom thought :-).
1285 (5) Because of (4), we can now allow for the use of $prvscheck_result
1286 inside the third argument.
1288 PH/03 For some reason, the default setting of PATH when running a command from
1289 a pipe transport was just "/usr/bin". I have changed it to
1292 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1293 anything to be listed in the output from -bV.
1295 PH/05 When a filter generated an autoreply, the entire To: header line was
1296 quoted in the delivery log line, like this:
1298 => >A.N.Other <ano@some.domain> <original@ddress> ...
1300 This has been changed so that it extracts the operative address. There
1301 may be more than one such address. If so, they are comma-separated, like
1304 => >ano@some.domain,ona@other.domain <original@ddress> ...
1306 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1307 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1308 not being looked up in the rDNS to get a host name, Exim was showing the
1309 IP address twice in Received: lines, even though the IP addresses were
1310 identical. For example:
1312 Received: from [1.2.3.4] (helo=[1.2.3.4])
1314 However, if the real host name was known, it was omitting the HELO data
1315 if it matched the actual IP address. This has been tidied up so that it
1316 doesn't show the same IP address twice.
1318 PH/07 When both +timestamp and +memory debugging was on, the value given by
1319 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1320 called by the debug printing, thereby overwriting the timestamp buffer.
1321 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1323 PH/08 When the original message was included in an autoreply transport, it
1324 always said "this is a copy of the message, including all the headers",
1325 even if body_only or headers_only was set. It now gives an appropriate
1328 PH/09 Applied a patch from the Sieve maintainer which:
1330 o fixes some comments
1331 o adds the (disabled) notify extension core
1332 o adds some debug output for the result of if/elsif tests
1333 o points to the current vacation draft in the documentation
1334 and documents the missing references header update
1338 o fixes a bug in processing the envelope test (when testing
1339 multiple envelope elements, the last element determinted the
1342 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1343 Electronic Mail") by including:
1345 Auto-submitted: auto-generated
1347 in the messages that it generates (bounce messages and others, such as
1348 warnings). In the case of bounce messages for non-SMTP mesages, there was
1349 also a typo: it was using "Auto_submitted" (underscore instead of
1350 hyphen). Since every message generated by Exim is necessarily in response
1351 to another message, thes have all been changed to:
1353 Auto-Submitted: auto-replied
1355 in accordance with these statements in the RFC:
1357 The auto-replied keyword:
1359 - SHOULD be used on messages sent in direct response to another
1360 message by an automatic process,
1362 - MUST NOT be used on manually-generated messages,
1364 - MAY be used on Delivery Status Notifications (DSNs) and Message
1365 Disposition Notifications (MDNs),
1367 - MUST NOT be used on messages generated by automatic or periodic
1368 processes, except for messages which are automatic responses to
1371 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1372 to the default Received: header definition.
1374 PH/12 Added log selector acl_warn_skipped (default on).
1376 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1377 variables because (a) they are in the wrong storage pool and (b) even if
1378 they were copied, it wouldn't work properly because of the caching.
1380 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1381 checking when decoding. Apparently there are clients that generate
1382 overlong encoded strings. Why am I not surprised?
1384 PH/15 If the first argument of "${if match_address" was not empty, but did not
1385 contain an "@" character, Exim crashed. Now it writes a panic log message
1386 and treats the condition as false.
1388 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1390 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1391 extension "envelope-auth". The code is finished and in agreement with
1392 other implementations, but there is no documentation so far and in fact,
1393 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1394 not changing the active code.
1396 Print executed "if" and "elsif" statements when debugging is used. This
1397 helps a great deal to understand what a filter does.
1399 Document more things not specified clearly in RFC3028. I had all this
1400 sorted out, when out of a sudden new issues came to my mind. Oops."
1402 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1405 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1406 canonical form (as documented). However, after a host name lookup from
1407 the IP address, check_host() was doing a simple string comparison with
1408 addresses acquired from the DNS when checking that the found name did
1409 have the original IP as one of its addresses. Since any found IPv6
1410 addresses are likely to be in abbreviated form, the comparison could
1411 fail. Luckily, there already exists a function for doing the comparison
1412 by converting both addresses to binary, so now that is used instead of
1413 the text comparison.
1415 PH/20 There was another similar case to PH/19, when a complete host name was
1416 given in a host list; looking up its IP address could give an abbreviated
1417 form, whereas the current host's name might or might not be abbreviated.
1418 The same fix has been applied.
1424 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1425 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1428 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1429 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1431 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1433 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1434 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1435 Also, for "Auto-Submitted", treat anything other than "no" as
1436 non-personal, in accordance with RFC 3834. (Previously it treated
1437 anything starting "auto-" as non-personal.)
1439 TF/01 The control=submission/name=... option had a problem with syntax
1440 errors if the name included a slash character. The /name= option
1441 now slurps the rest of the string, so it can include any characters
1442 but it must come last in the list of options (after /sender_retain
1445 PH/05 Some modifications to the interface to the fake nameserver for the new
1453 TK/01 Added the "success_on_redirect" address verification option. See
1454 NewStuff for rationale and an example.
1456 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1458 PH/02 Patch to exigrep to allow it to work on syslog lines.
1460 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1461 fread() to skip over the body file's header line, because in Cygwin the
1462 header line is locked and is inaccessible.
1464 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1465 co-exist for some time) to make it clear that it is the Exim ID that is
1466 referenced, not the Message-ID: header line.
1468 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1469 string_format() function, because snprintf() does not exist on all
1472 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1473 ${stat: expansion item.
1475 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1476 protocol synchronization error", to keep the pedants happy.
1478 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1479 well as for IRIX systems, when gcc is being used. See the host.c source
1482 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1484 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1487 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1488 transport and to the smtp transport in LMTP mode.
1490 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1492 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1493 run a filter in a subprocess. This could lead to confusion in subsequent
1494 lookups in the parent process. There should also be a search_tidyup() at
1495 the end of the subprocess.
1497 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1498 only if the host matched helo_try_verify_hosts, which caused the
1499 verification to occur when the EHLO/HELO command was issued. The ACL just
1500 tested the remembered result. Now, if a previous verification attempt has
1501 not happened, "verify = helo" does it there and then.
1503 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1505 TK/03 Fix log output including CR from clamd.
1507 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1508 exist provoked a memory error which could cause a segfault.
1510 PH/15 Installed PCRE 6.2
1512 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1514 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1515 of the problem. Specifically, suggested +O2 rather than +O1 for the
1518 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1520 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1521 record, the port was not correctly logged when the outgoing_port log
1522 selector was set (it logged the transort's default port).
1524 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1525 fallback_hosts, and "hosts" in the smtp transport.
1527 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1528 host errors such as "Connection refused".
1530 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1531 authentication with radiusclient 0.4.9:
1533 - Error returned from rc_read_config was caught wrongly
1534 - Username/password not passed on to radius server due to wrong length.
1536 The presumption is that some radiusclient API changes for 4.51/PH/17
1537 were not taken care of correctly. The code is still untested by me (my
1538 Linux distribution still has 0.3.2 of radiusclient), but it was
1539 contributed by a Radius user.
1541 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1542 expanding the "port" option of the smtp transport.
1544 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1545 while reading a MIME header. Thanks to Tom Hughes for a patch.
1547 PH/24 Include config.h inside local_scan.h so that configuration settings are
1550 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1551 that in an ACL for RCPT (for example), you can examine exactly what was
1554 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1555 commands, but it was not correctly comparing the address with the actual
1556 client host address. Thus, it would show the EHLO address in Received:
1557 header lines when this was not necessary.
1559 PH/27 Added the % operator to ${eval:}.
1561 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1562 it should be ignoring failures (because with -C, for example, it has lost
1563 privilege). It wasn't ignoring creation failures other than "already
1566 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1569 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1570 because an input line was too long, either on its own, or by virtue of
1571 too many continuations, the temporary file was not being removed, and the
1572 return code was incorrect.
1574 PH/31 Missing "BOOL" in function definition in filtertest.c.
1576 PH/32 Applied Sieve patches from the maintainer.
1578 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1580 PH/33 Added "verify = not_blind".
1582 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1583 Local/Makefile (with some defaults set). These are used in built scripts
1584 such as exicyclog, but they have never been used in the exim_install
1585 script (though there are many overriding facilities there). I have
1586 arranged that the exim_install script now takes note of these two
1589 PH/35 Installed configuration files for Dragonfly.
1591 PH/36 When a locally submitted message by a trusted user did not contain a
1592 From: header, and the sender address was obtained from -f or from an SMTP
1593 MAIL command, and the trusted user did not use -F to supply a sender
1594 name, $originator_name was incorrectly used when constructing a From:
1595 header. Furthermore, $originator_name was used for submission mode
1596 messages from external hosts without From: headers in a similar way,
1597 which is clearly wrong.
1599 PH/37 Added control=suppress_local_fixups.
1601 PH/38 When log_selector = +received_sender was set, and the addition of the
1602 sender made the log line's construction buffer exactly full, or one byte
1603 less than full, an overflow happened when the terminating "\n" was
1606 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1607 when the result of a list match is failure because a DNS lookup failed.
1609 PH/40 RM_COMMAND is now used in the building process.
1611 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1612 the "build-* directories that it finds.
1614 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1615 address in a domain literal was a prefix of an interface address.
1617 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1618 when verifying a sender address, unless rewrite_headers is false.
1620 PH/44 Wrote a long comment about why errors_to addresses are verified as
1621 recipients, not senders.
1623 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1624 the ratelimit ACL was added.
1626 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1628 PH/46 Added extra information about PostgreSQL errors to the error string.
1630 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1631 avoiding the need to install special zones in a real server. This is
1632 backwards compatible; if it can't find the fake resolver, it drops back.
1633 Thus, both old and new test suites can be run.
1635 TF/02 Added util/ratelimit.pl
1637 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1638 clock is set back in time.
1640 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1641 Candler <B.Candler@pobox.com>.
1643 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1644 OK for addresses that are the result of redirections.
1646 PH/48 A number of further additions for the benefit of the new test suite,
1647 including a fake gethostbyname() that interfaces to the fake DNS resolver
1650 TF/06 The fix for widen_domains has also been applied to qualify_single and
1651 search_parents which are the other dnslookup options that can cause
1654 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1657 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1659 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1660 service" if it encountered an MX record with an empty target hostname.
1661 The message is now "an MX or SRV record indicated no SMTP service".
1663 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1664 if the DNS of the sending site is misconfigured. This is quite a
1665 common situation. This change restores the behaviour of treating a
1666 helo verification defer as a failure.
1668 PH/51 If self=fail was set on a router, the bounce message did not include the
1669 actual error message.
1675 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1677 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1678 command itself ended in error, the underlying message about the transport
1679 filter timeout was being overwritten with the pipe command error. Now the
1680 underlying error message should be appended to the second error message.
1682 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1684 PH/02 Reduce the amount of output that "make" produces by default. Full output
1685 can still be requested.
1687 PH/03 The warning log line about a condition test deferring for a "warn" verb
1688 was being output only once per connection, rather than after each
1689 occurrence (because it was using the same function as for successful
1690 "warn" verbs). This seems wrong, so I have changed it.
1692 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1693 it should not have, which might have caused a crash in the right
1694 circumstances, but probably never did.
1696 PH/04 Installed a modified version of Tony Finch's patch to make submission
1697 mode fix the return path as well as the Sender: header line, and to
1698 add a /name= option so that you can make the user's friendly name appear
1701 TF/03 Added the control = fakedefer ACL modifier.
1703 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1704 Mark Lowes for thorough testing.
1706 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1708 TK/03 Merged latest SRS patch from Miles Wilton.
1710 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1711 with the definition in sysexits.h (which is #included earlier).
1712 Fortunately, Exim does not actually use EX_OK. The code used to try to
1713 preserve the sysexits.h value, by assumimg that macro definitions were
1714 scanned for macro replacements. I have been disabused of this notion,
1715 so now the code just undefines EX_OK before #including unistd.h.
1717 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1718 in the smtp transport. When a block could not be written in a single
1719 write() function, the timeout was being re-applied to each part-write.
1720 This seems wrong - if the receiver was accepting one byte at a time it
1721 would take for ever. The timeout is now adjusted when this happens. It
1722 doesn't have to be particularly precise.
1724 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1725 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1727 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1728 without a domain if the check to <postmaster@domain> fails.
1730 SC/01 Eximstats: added -xls and the ability to specify output files
1731 (patch written by Frank Heydlauf).
1733 SC/02 Eximstats: use FileHandles for outputing results.
1735 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1737 SC/04 Eximstats: fixed display of large numbers with -nvr option
1739 SC/05 Eximstats: fixed merging of reports with empty tables.
1741 SC/06 Eximstats: added the -include_original_destination flag
1743 SC/07 Eximstats: removed tabs and trailing whitespace.
1745 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1747 TK/06 MBOX spool code: Add real "From " MBOX separator line
1748 so the .eml file is really in mbox format (even though
1749 most programs do not really care). Patch from Alex Miller.
1751 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1752 The latter is generated from $received_to and is only set if the
1753 message has one envelope recipient. SA can use these headers,
1754 obviously out-of-the-box. Patch from Alex Miller.
1756 PH/08 The ${def test on a variable was returning false if the variable's
1757 value was "0", contrary to what the specification has always said!
1758 The result should be true unless the variable is empty.
1760 PH/09 The syntax error of a character other than { following "${if
1761 def:variable_name" (after optional whitespace) was not being diagnosed.
1762 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1763 accidental colon was present, for example, could give incorrect results.
1765 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1766 result is used (not including appendfile, where other changes are about
1769 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1770 This involved changing a lot of size variables from int to off_t. It
1771 should work with maildirs and everything.
1773 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1774 spamd dying while we are connected to it.
1776 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1779 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1780 function for BDB 4.3.
1782 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1784 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1785 the sources for PCRE have been re-arranged and more files are now
1788 PH/15 The code I had for printing potentially long long variables in PH/11
1789 above was not the best (it lost precision). The length of off_t variables
1790 is now inspected at build time, and an appropriate printing format (%ld
1791 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1792 to be "long long int" or "long int". This is needed for the internal
1793 formatting function string_vformat().
1795 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1796 the configuration file to be ":syslog", then the script "guesses" where
1797 the logs files are, rather than using the compiled in default. In our
1798 case the guess is not the same as the compiled default, so the script
1799 suddenly stopped working when I started to use syslog. The patch checks
1800 to see if log_file_path is "". If so, it attempts to read it from exim
1801 with no configuration file to get the compiled in version, before it
1802 falls back to the previous guessing code."
1804 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1805 implementing BATV in an Exim configuration. See NewStuff for the gory
1808 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1809 Makefile that are specific to HP-UX.
1811 PH/18 If the "use_postmaster" option was set for a recipient callout together
1812 with the "random" option, the postmaster address was used as the MAIL
1813 FROM address for the random test, but not for the subsequent recipient
1814 test. It is now used for both.
1816 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1817 patch removes a few documentation additions to RFC 3028, because the
1818 latest draft now contains them. It adds the new en;ascii-case comparator
1819 and a new error check for 8bit text in MIME parts. Comparator and
1820 require names are now matched exactly. I enabled the subaddress
1821 extension, but it is not well tested yet (read: it works for me)."
1823 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1824 rework some of the code of TK/09 above to avoid the hardwired use of
1825 "%lld" and "long long". Replaced the call to snprintf() with a call to
1828 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1829 records point to non-existent hosts", "retry timeout exceeded", and
1830 "retry time not reached for any host after a long failure period".
1832 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1833 experimental DomainKeys support:
1835 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1836 (2) On an error such as an illegally used "control", the wrong name for
1837 the control was given.
1839 These problems did NOT occur unless DomainKeys support was compiled.
1841 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1843 PH/24 Added ${if match_ip condition.
1845 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1846 databases so that it will be absolutely obvious if a crash occurs in the
1847 DB library. This is a regular occurrence (often caused by mis-matched
1850 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1851 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1852 picked up on a user's system that detects such things. There doesn't seem
1853 to be a gcc warning option for this - only an attribute that has to be
1854 put on the function's prototype. It seems that in Fedora Core 4 they have
1855 set this on a number of new functions. No doubt there will be more in due
1858 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1859 specify a transport. However, if an address that was verified by such a
1860 router was the subject of a callout, Exim crashed because it tried to
1861 read the rcpt_include_affixes from the non-existent transport. Now it
1862 just assumes that the setting of that option is false. This bug was
1863 introduced by 4.51/PH/31.
1865 PH/28 Changed -d+all to exclude +memory, because that information is very
1866 rarely of interest, but it makes the output a lot bigger. People tend to
1867 do -d+all out of habit.
1869 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1870 code in os-type was giving problems when libc.so lives in lib64, like on
1873 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1874 aren't the modern standard, and it seems that some systems' include files
1875 don't always have them. Exim was already checking for some of the newer
1876 ones like T_AAAA, and defining it itself. I've added checks for all the
1877 record types that Exim uses.
1879 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1880 not automatically generating a new one, as it is supposed to. This
1881 prevented TLS from working. If the file did exist, but contained invalid
1882 data, a new version was generated, as expected. It was only the case of a
1883 non-existent file that was broken.
1885 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1886 with a change in libdomainkeys > 0.64.
1888 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1889 from DNS. If the selector record carries the flag, it now has
1890 precedence over the domain-wide flag.
1892 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1894 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1895 the use of a port name that isn't defined in /etc/services) occurred, the
1896 message was deferred as in a normal delivery, and thus remained on the
1897 spool, instead of being failed because of the mua_wrapper setting. This
1898 is now fixed, and I tidied up some of the mua_wrapper messages at the
1901 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1902 the messages in a hash of arrays rather than using individual hashes.
1903 This is a bit cleaner and results in dramatic memory savings, albeit
1904 at a slight CPU cost.
1906 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1907 as requested by Marc Sherman.
1909 SC/10 Eximstats: added histograms for user specified patterns as requested
1912 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1914 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1915 fopen() in the content-scanning modules that did not already have it.
1921 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1922 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1924 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1926 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1928 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1929 to test Sieve filters that use "vacation".
1931 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1932 that changes the way the GnuTLS parameters are stored in the cache file.
1933 The new format can be generated externally. For backward compatibility,
1934 if the data in the cache doesn't make sense, Exim assumes it has read an
1935 old-format file, and it generates new data and writes a new file. This
1936 means that you can't go back to an older release without removing the
1939 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1940 work if there are any delivery delays because "one_time" forces the
1941 parent to be marked "delivered", so its unseen clone is never tried
1942 again. For this reason, Exim now forbids the simultaneous setting of
1945 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1946 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1947 entry if you want to know the details. The fix, however, neglected to
1948 consider the case when local delivery batching is involved. The test for
1949 "previously delivered" was not happening when checking to see if an
1950 address could be batched with a previous (undelivered) one; under
1951 certain circumstances this could lead to multiple deliveries to the same
1954 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1955 in its include files, and this causes problems building Exim.
1957 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1958 header_syntax) but Exim was just ignoring anything given after a slash.
1959 In particular, this caused confusion with an attempt to use "verify =
1960 reverse_host_lookup/defer_ok". An error is now given when options are
1961 supplied for verify items that do not have them. (Maybe reverse_host_
1962 lookup should have a defer_ok option, but that's a different point.)
1964 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1965 defined by RFC 821) to 2048, because there were problems with some AUTH
1966 commands, and RFC 1869 says the size should be increased for extended
1967 SMTP commands that take arguments.
1969 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1972 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1973 "unknown" error; now it says that the functionality isn't in the binary.
1975 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1976 an address' error message when a string expansion fails (syntax or
1977 whatever). Otherwise the password may appear in the log. Following change
1978 PH/42 below, there is no longer a chance of it appearing in a bounce
1981 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1983 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1984 of its addresses was ever tried. (Bugzilla bug #2.)
1986 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1987 the result incorrectly in the debug output. (It correctly added a newline
1988 to what was transported.)
1990 TF/01 Added $received_time.
1992 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1993 commented out examples of how to interface to a virus scanner and to
1994 SpamAssassin. Also added commented examples of av_scanner and
1995 spamd_address settings.
1997 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1998 and controls are allowed in which ACLs. There were a couple of minor
1999 errors. Some of the entries in the conditions table (which is a table of
2000 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2001 negation of where the condition IS allowed.
2003 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2005 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2006 header file does not have a version number, so I've had to invent a new
2007 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2008 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2009 radiusclient), but it was contributed by a Radius user.
2011 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2012 files or directories, for OpenSSL.
2014 PH/19 When an Exim process that is running as root has to create an Exim log
2015 file, it does so in a subprocess that runs as exim:exim so as to get the
2016 ownership right at creation (otherwise, other Exim processes might see
2017 the file with the wrong ownership). There was no test for failure of this
2018 fork() call, which would lead to the process getting stuck as it waited
2019 for a non-existent subprocess. Forks do occasionally fail when resources
2020 run out. I reviewed all the other calls to fork(); they all seem to check
2023 PH/20 When checking for unexpected SMTP input at connect time (before writing
2024 the banner), Exim was not dealing correctly with a non-positive return
2025 from the read() function. If the client had disconnected by this time,
2026 the result was a log entry for a synchronization error with an empty
2027 string after "input=" when read() returned zero. If read() returned -1
2028 (an event I could not check), uninitialized data bytes were printed.
2029 There were reports of junk text (parts of files, etc) appearing after
2032 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2034 PH/22 Added support for macro redefinition, and (re)definition in between
2035 driver and ACL definitions.
2037 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2038 forgetting to use the resulting value; it was using the unexpanded value.
2040 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2041 hadn't been configured. The fix is from Juergen Kreileder, who
2042 understands it better than I do:
2044 "Here's what I see happening with three configured cyrus_sasl
2045 authenticators configured (plain, login, cram-md5):
2047 On startup auth_cyrus_sasl_init() gets called for each of these.
2048 This means three calls to sasl_listmech() without a specified mech_list.
2049 => SASL tests which mechs of all available mechs actually work
2050 => three warnings about OTP not working
2051 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2053 With the patch, sasl_listmech() also gets called three times. But now
2054 SASL's mech_list option is set to the server_mech specified in the the
2055 authenticator. Or in other words, the answer from sasl_listmech()
2056 gets limited to just the mech you're testing for (which is different
2058 => the return list contains just 'plain' or 'login', 'cram-md5' or
2059 nothing depending on the value of ob->server_mech.
2061 I've just tested the patch: Authentication still works fine,
2062 unavailable mechs specified in the exim configuration are still
2063 caught, and the auth.log warnings about OTP are gone."
2065 PH/25 When debugging is enabled, the contents of the command line are added
2066 to the debugging output, even when log_selector=+arguments is not
2069 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2070 answer is "GNU", and only if the return is "GNU/something" is the answer
2073 PH/27 $acl_verify_message is now set immediately after the failure of a
2074 verification in an ACL, and so is available in subsequent modifiers. In
2075 particular, the message can be preserved by coding like this:
2077 warn !verify = sender
2078 set acl_m0 = $acl_verify_message
2080 Previously, $acl_verify_message was set only while expanding "message"
2081 and "log_message" when a very denied access.
2083 PH/28 Modified OS/os.c-Linux with
2085 -#ifndef OS_LOAD_AVERAGE
2086 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2088 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2089 nomenclature these days.)
2091 PH/29 Installed patch from the Sieve maintainer that adds the options
2092 sieve_useraddress and sieve_subaddress to the redirect router.
2094 PH/30 In these circumstances:
2095 . Two addresses routed to the same list of hosts;
2096 . First host does not offer TLS;
2097 . First host accepts first address;
2098 . First host gives temporary error to second address;
2099 . Second host offers TLS and a TLS session is established;
2100 . Second host accepts second address.
2101 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2102 and peerdn, if requested) that were in fact used only for the second
2105 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2106 attention to any local part prefix or suffix that was matched by the
2107 router that accepted the address. It now behaves in the same way as it
2108 does for delivery: the affixes are removed from the local part unless
2109 rcpt_include_affixes is set on the transport.
2111 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2112 timeout during the DATA phase of an incoming message.
2114 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2115 applied a patch sanctioned by the Sieve maintainer.
2117 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2118 the uid or gid is negative. A case of a negative gid caused this to be
2119 noticed. The fix allows for either to be negative.
2121 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2122 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2125 PH/36 The change PH/12 above was broken. Fixed it.
2127 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2128 the grounds that routing the same address twice would always produce the
2129 same answer. This might have been true once, but it is certainly no
2130 longer true now. Routing a child address may depend on the previous
2131 routing that produced that child. Some complicated redirection strategies
2132 went wrong when messages had multiple recipients, and made Exim's
2133 behaviour dependent on the order in which the addresses were given.
2135 I have moved the duplicate checking until after the routing is complete.
2136 Exim scans the addresses that are assigned to local and remote
2137 transports, and removes any duplicates. This means that more work will be
2138 done, as duplicates will always all be routed, but duplicates are
2139 presumably rare, so I don't expect this is of any significance.
2141 For deliveries to pipes, files, and autoreplies, the duplicate checking
2142 still happens during the routing process, since they are not going to be
2145 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2146 It corrects a timeout issue with spamd. This is Ian's comment: "The
2147 background is that sometimes spamd either never reads data from a
2148 connection it has accepted, or it never writes response data. The exiscan
2149 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2150 blindly assumes that writes won't block so it may never time out."
2152 PH/39 Allow G after quota size as well as K and M.
2154 PH/40 The value set for $authenticated_id in an authenticator may not contain
2155 binary zeroes or newlines because the value is written to log lines and
2156 to spool files. There was no check on this. Now the value is run through
2157 the string_printing() function so that such characters are converted to
2158 printable escape sequences.
2160 PH/41 $message_linecount is a new variable that contains the total number of
2161 lines in the message. Compare $body_linecount, which is the count for the
2164 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2165 bounce and delay warning messages, except in certain special cases, which
2168 (a) An SMTP error message from a remote host;
2169 (b) A message specified in a :fail: redirection;
2170 (c) A message specified in a "fail" command in a system filter;
2171 (d) A message specified in a FAIL return from the queryprogram router;
2172 (e) A message specified by the cannot_route_message router option.
2174 In these cases only, Exim does include the error details in bounce and
2175 warning messages. There are also a few cases where bland messages such
2176 as "unrouteable address" or "local delivery error" are given.
2178 PH/43 $value is now also set for the "else" part of a ${run expansion.
2180 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2181 being worked on, but at least Exim now implements the latest version to
2184 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2185 process to complete was treated as a delivery failure, a timeout while
2186 writing the message to the pipe was logged, but erroneously treated as a
2187 successful delivery. Such timeouts include transport filter timeouts. For
2188 consistency with the overall process timeout, these timeouts are now
2189 treated as errors, giving rise to delivery failures by default. However,
2190 there is now a new Boolean option for the pipe transport called
2191 timeout_defer, which, if set TRUE, converts the failures into defers for
2192 both kinds of timeout. A transport filter timeout is now identified in
2195 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2196 systems where "make" and "gmake" are different, calling "gmake" at top
2197 level broke things. I've arranged for the value of $(MAKE) to be passed
2198 from the Makefile to this script so that it can call the same version of
2202 A note about Exim versions 4.44 and 4.50
2203 ----------------------------------------
2205 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2206 changes of various kinds. As a consequence, a big documentation update was
2207 needed. This delayed the release for rather longer than seemed good, especially
2208 in the light of a couple of (minor) security issues. Therefore, the changes
2209 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2210 release. So 4.44 and 4.50 are in effect two different branches that both start
2213 I have left the 4.50 change log unchanged; it contains all the changes since
2214 4.43. The change log for 4.44 is below; many of its items are identical to
2215 those for 4.50. This seems to be the most sensible way to preserve the
2216 historical information.
2222 1. Minor wording change to the doc/README.SIEVE file.
2224 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2225 computation of the current number of files was incorrect.
2227 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2228 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2231 4. Give more explanation in the error message when the command for a transport
2232 filter fails to execute.
2234 5. There are several places where Exim runs a non-Exim command in a
2235 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2236 was being done only for the command run by the queryprogram router. It is
2237 now done for all such subprocesses. The other cases are: ${run, transport
2238 filters, and the commands run by the lmtp and pipe transports.
2240 6. Added CONFIGURE_GROUP build-time option.
2242 7. Some older OS have a limit of 256 on the maximum number of file
2243 descriptors. Exim was using setrlimit() to set 1000 as a large value
2244 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2245 systems. I've change it so that if it can't get 1000, it tries for 256.
2247 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2248 was an oversight, and furthermore, ever since the addition of extra
2249 controls (e.g. 4.43/32), the checks on when to allow different forms of
2250 "control" were broken. There should now be diagnostics for all cases when a
2251 control that does not make sense is encountered.
2253 9. Added the /retain_sender option to "control=submission".
2255 10. $recipients is now available in the predata ACL (oversight).
2257 11. Tidy the search cache before the fork to do a delivery from a message
2258 received from the command line. Otherwise the child will trigger a lookup
2259 failure and thereby defer the delivery if it tries to use (for example) a
2260 cached ldap connection that the parent has called unbind on.
2262 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2263 of $address_data from the recipient verification was clobbered by the
2264 sender verification.
2266 13. The value of address_data from a sender verification is now available in
2267 $sender_address_data in subsequent conditions in the ACL statement.
2269 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2271 15. Added a new option "connect=<time>" to callout options, to set a different
2274 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2275 was its contents. (It was OK if the option was not defined at all.)
2277 17. A "Completed" log line is now written for messages that are removed from
2278 the spool by the -Mrm option.
2280 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2281 information about exactly what failed.
2283 19. Added -dd to debug only the daemon process.
2285 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2286 handles timeouts, both on the server side and network timeouts. Renamed the
2287 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2289 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2290 It is now set to "smtps".
2292 22. $host_address is now set to the target address during the checking of
2293 ignore_target_hosts.
2295 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2296 being passed; this would have caused $sender_host_name to have been used if
2297 matching the list had actually called for a host name (not very likely,
2298 since this list is usually IP addresses). A host name is now passed as
2301 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2302 code that specifies a non-restarting handler (typically sigaction() in
2303 modern systems) in an attempt to fix a rare and obscure crash bug.
2305 25. Narrowed the window for a race in the daemon that could cause it to ignore
2306 SIGCHLD signals. This is not a major problem, because they are used only to
2307 wake it up if nothing else does.
2309 26. A malformed maildirsize file could cause Exim to calculate negative values
2310 for the mailbox size or file count. Odd effects could occur as a result.
2311 The maildirsize information is now recalculated if the size or filecount
2314 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2315 support for a long time. Removed HAVE_SYS_VFS_H.
2317 28. Installed the latest version of exipick from John Jetmore.
2319 29. In an address list, if the pattern was not a regular expression, an empty
2320 subject address (from a bounce message) matched only if the pattern was an
2321 empty string. Non-empty patterns were not even tested. This was the wrong
2322 because it is perfectly reasonable to use an empty address as part of a
2323 database query. An empty address is now tested by patterns that are
2324 lookups. However, all the other forms of pattern expect the subject to
2325 contain a local part and a domain, and therefore, for them, an empty
2326 address still always fails if the pattern is not itself empty.
2328 30. Exim went into a mad DNS loop when attempting to do a callout where the
2329 host was specified on an smtp transport, and looking it up yielded more
2330 than one IP address.
2332 31. Re-factored the code for checking spool and log partition space into a
2333 function that finds that data and another that does the check. The former
2334 is then used to implement four new variables: $spool_space, $log_space,
2335 $spool_inodes, and $log_inodes.
2337 32. The RFC2047 encoding function was originally intended for short strings
2338 such as real names; it was not keeping to the 75-character limit for
2339 encoded words that the RFC imposes. It now respects the limit, and
2340 generates multiple encoded words if necessary. To be on the safe side, I
2341 have increased the buffer size for the ${rfc2047: expansion operator from
2344 33. It is now permitted to omit both strings after an "if" condition; if the
2345 condition is true, the result is "true". As before, when the second string
2346 is omitted, a false condition yields an empty string. This makes it less
2347 cumbersome to write custom ACL and router conditions.
2349 34. Failure to deliver a bounce message always caused it to be frozen, even if
2350 there was an errors_to setting on the router. The errors_to setting is now
2353 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2354 canonical form (fully expanded) before being placed in
2355 $sender_host_address.
2357 36. The table in the code that translates DNS record types into text (T_A to
2358 "A" for instance) was missing entries for NS and CNAME. It is just possible
2359 that this could have caused confusion if both these types were looked up
2360 for the same domain, because the text type is used as part of Exim's
2361 per-process caching. But the chance of anyone hitting this buglet seems
2364 37. The dnsdb lookup has been extended in a number of ways.
2366 (1) There is a new type, "zns", which walks up the domain tree until it
2367 finds some nameserver records. It should be used with care.
2369 (2) There is a new type, "mxh", which is like "mx" except that it returns
2370 just the host names, not the priorities.
2372 (3) It is now possible to give a list of domains (or IP addresses) to be
2373 looked up. The behaviour when one of the lookups defers can be
2374 controlled by a keyword.
2376 (4) It is now possible to specify the separator character for use when
2377 multiple records are returned.
2379 38. The dnslists ACL condition has been extended: it is now possible to supply
2380 a list of IP addresses and/or domains to be looked up in a particular DNS
2383 39. Added log_selector=+queue_time_overall.
2385 40. When running the queue in the test harness, wait just a tad after forking a
2386 delivery process, to get repeatability of debugging output.
2388 41. Include certificate and key file names in error message when GnuTLS fails
2389 to set them up, because the GnuTLS error message doesn't include the name
2390 of the failing file when there is a problem reading it.
2392 42. Allow both -bf and -bF in the same test run.
2394 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2396 44. The "Exiscan patch" is now merged into the mainline Exim source.
2398 45. Sometimes the final signoff response after QUIT could fail to get
2399 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2400 < 0 before doing a fflush(). This bug looks as though it goes back to the
2401 introduction of TLS in release 3.20, but "sometimes" must have been rare
2402 because the tests only now provoked it.
2404 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2405 (this can affect the format of dates).
2407 47. exim_tidydb, when checking for the continued existence of a message for
2408 which it has found a message-specific retry record, was not finding
2409 messages that were in split spool directories. Consequently, it was
2410 deleting retry records that should have stayed in existence.
2412 48. Steve fixed some bugs in eximstats.
2414 49. The SPA authentication driver was not abandoning authentication and moving
2415 on to the next authenticator when an expansion was forced to fail,
2416 contradicting the general specification for all authenticators. Instead it
2417 was generating a temporary error. It now behaves as specified.
2419 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2420 (the order specifies the preference for clients). The order is now AES256,
2421 AES128, 3DES, ARCFOUR128.
2423 51. Small patch to Sieve code - explicitly set From: when generating an
2426 52. Exim crashed if a remote delivery caused a very long error message to be
2427 recorded - for instance if somebody sent an entire SpamAssassin report back
2428 as a large number of 550 error lines. This bug was coincidentally fixed by
2429 increasing the size of one of Exim's internal buffers (big_buffer) that
2430 happened as part of the Exiscan merge. However, to be on the safe side, I
2431 have made the code more robust (and fixed the comments that describe what
2434 53. Now that there can be additional text after "Completed" in log lines (if
2435 the queue_time_overall log selector is set), a one-byte patch to exigrep
2436 was needed to allow it to recognize "Completed" as not the last thing in
2439 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2440 patch that reportedly fixes this has been added. I am not expert enough to
2441 create a test for it. This is what the patch creator wrote:
2443 "I found a little strange behaviour of ldap code when working with
2444 Windows 2003 AD Domain, where users was placed in more than one
2445 Organization Units. When I tried to give exim partial DN, the exit code
2446 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2447 But simultaneously result of request was absolutely normal ldap result,
2448 so I produce this patch..."
2450 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2451 REFERENCE, so I have modified the code to exclude the patch when that macro
2454 55. Some experimental protocols are using DNS PTR records for new purposes. The
2455 keys for these records are domain names, not reversed IP addresses. The
2456 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2457 leaves it alone. Component reversal etc. now happens only for IP addresses.
2460 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2462 57. Double the size of the debug message buffer (to 2048) so that more of very
2463 long debug lines gets shown.
2465 58. The exicyclog utility now does better if the number of log files to keep
2466 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2468 59. Two changes related to the smtp_active_hostname option:
2470 (1) $smtp_active_hostname is now available as a variable.
2471 (2) The default for smtp_banner uses $smtp_active_hostname instead
2472 of $primary_hostname.
2474 60. The host_aton() function is supposed to be passed a string that is known
2475 to be a valid IP address. However, in the case of IPv6 addresses, it was
2476 not checking this. This is a hostage to fortune. Exim now panics and dies
2477 if the condition is not met. A case was found where this could be provoked
2478 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2479 components; fortuitously, this particular loophole had already been fixed
2480 by change 4.50/55 above.
2482 If there are any other similar loopholes, the new check in host_aton()
2483 itself should stop them being exploited. The report I received stated that
2484 data on the command line could provoke the exploit when Exim was running as
2485 exim, but did not say which command line option was involved. All I could
2486 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2487 running as the user.
2490 61. There was a buffer overflow vulnerability in the SPA authentication code
2491 (which came originally from the Samba project). I have added a test to the
2492 spa_base64_to_bits() function which I hope fixes it.
2495 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2496 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2498 63. The daemon start-up calls getloadavg() while still root for those OS that
2499 need the first call to be done as root, but it missed one case: when
2500 deliver_queue_load_max is set with deliver_drop_privilege. This is
2501 necessary for the benefit of the queue runner, because there is no re-exec
2502 when deliver_drop_privilege is set.
2504 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2505 This has been fixed.
2507 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2508 was in use, was not putting the data itself into the right store pool;
2509 consequently, it could be overwritten for a subsequent message in the same
2510 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2513 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2515 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2516 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2517 the function were treating the return as a boolean value, which happened to
2518 work because 0=false and not-0=true, but is not correct code.
2520 68. The host_aton() function was not handling scoped IPv6 addresses (those
2521 with, for example, "%eth0" on the end) correctly.
2523 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2524 negated items (that is, ~something) in unsigned ints. Some compilers
2525 apparently mutter when there is no cast.
2527 70. If an address verification called from an ACL failed, and did not produce a
2528 user-specific message (i.e. there was only a "system" message), nothing was
2529 put in $acl_verify_message. In this situation, it now puts the system
2532 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2533 session; change 4.31/43 added the unwanted input to the log line - except
2534 that it did not do this in the start of session case. It now does.
2536 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2537 This is wrong and can cause the other end to generate a synchronization
2538 error if it is another Exim or anything else that does the synchronization
2539 check. A QUIT command is no longer sent after a timeout.
2541 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2542 during host lookups.
2544 74. The defer_ok option of callout verification was not working if it was used
2545 when verifying addresses in header lines, that is, for this case:
2547 verify = header_sender/callout=defer_ok
2549 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2550 those file descriptors could be used for SMTP connections. If anything
2551 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2552 could be sent to the SMTP client, causing chaos. The daemon now opens
2553 stdin, stdout, and stderr to /dev/null when it puts itself into the
2556 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2557 log by default. The user can override this with suitable Perl magic.
2559 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2560 the log message when discard triggers, was not working for the DATA ACL or
2561 for the non-SMTP ACL.
2563 78. Error message wording change in sieve.c.
2565 79. If smtp_accept_max_per_host was set, the number of connections could be
2566 restricted to fewer than expected, because the daemon was trying to set up
2567 a new connection before checking whether the processes handling previous
2568 connections had finished. The check for completed processes is now done
2569 earlier. On busy systems, this bug wouldn't be noticed because something
2570 else would have woken the daemon, and it would have reaped the completed
2573 80. If a message was submitted locally by a user whose login name contained one
2574 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2575 It caused a spool format error. I have fixed the spool reading code. A
2576 related problem was that the "from" clause in the Received: line became
2577 illegal because of the space(s). It is now covered by ${quote_local_part.
2579 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2582 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2584 83. Patch from Sieve maintainer to fix unterminated string problem in
2585 "vacation" handling.
2587 84. Some minor changes to the Linux configuration files to help with other
2588 OS variants using glibc.
2590 85. One more patch for Sieve to update vacation handling to latest spec.
2593 ----------------------------------------------------
2594 See the note above about the 4.44 and 4.50 releases.
2595 ----------------------------------------------------
2601 1. Change 4.43/35 introduced a bug that caused file counts to be
2602 incorrectly computed when quota_filecount was set in an appendfile
2605 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2606 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2609 3. Give more explanation in the error message when the command for a transport
2610 filter fails to execute.
2612 4. There are several places where Exim runs a non-Exim command in a
2613 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2614 was being done only for the command run by the queryprogram router. It is
2615 now done for all such subprocesses. The other cases are: ${run, transport
2616 filters, and the commands run by the lmtp and pipe transports.
2618 5. Some older OS have a limit of 256 on the maximum number of file
2619 descriptors. Exim was using setrlimit() to set 1000 as a large value
2620 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2621 systems. I've change it so that if it can't get 1000, it tries for 256.
2623 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2624 was an oversight, and furthermore, ever since the addition of extra
2625 controls (e.g. 4.43/32), the checks on when to allow different forms of
2626 "control" were broken. There should now be diagnostics for all cases when a
2627 control that does not make sense is encountered.
2629 7. $recipients is now available in the predata ACL (oversight).
2631 8. Tidy the search cache before the fork to do a delivery from a message
2632 received from the command line. Otherwise the child will trigger a lookup
2633 failure and thereby defer the delivery if it tries to use (for example) a
2634 cached ldap connection that the parent has called unbind on.
2636 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2637 of $address_data from the recipient verification was clobbered by the
2638 sender verification.
2640 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2641 was its contents. (It was OK if the option was not defined at all.)
2643 11. A "Completed" log line is now written for messages that are removed from
2644 the spool by the -Mrm option.
2646 12. $host_address is now set to the target address during the checking of
2647 ignore_target_hosts.
2649 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2650 being passed; this would have caused $sender_host_name to have been used if
2651 matching the list had actually called for a host name (not very likely,
2652 since this list is usually IP addresses). A host name is now passed as
2655 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2656 code that specifies a non-restarting handler (typically sigaction() in
2657 modern systems) in an attempt to fix a rare and obscure crash bug.
2659 15. Narrowed the window for a race in the daemon that could cause it to ignore
2660 SIGCHLD signals. This is not a major problem, because they are used only to
2661 wake it up if nothing else does.
2663 16. A malformed maildirsize file could cause Exim to calculate negative values
2664 for the mailbox size or file count. Odd effects could occur as a result.
2665 The maildirsize information is now recalculated if the size or filecount
2668 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2669 support for a long time. Removed HAVE_SYS_VFS_H.
2671 18. Updated exipick to current release from John Jetmore.
2673 19. Allow an empty sender to be matched against a lookup in an address list.
2674 Previously the only cases considered were a regular expression, or an
2677 20. Exim went into a mad DNS lookup loop when doing a callout where the
2678 host was specified on the transport, if the DNS lookup yielded more than
2681 21. The RFC2047 encoding function was originally intended for short strings
2682 such as real names; it was not keeping to the 75-character limit for
2683 encoded words that the RFC imposes. It now respects the limit, and
2684 generates multiple encoded words if necessary. To be on the safe side, I
2685 have increased the buffer size for the ${rfc2047: expansion operator from
2688 22. Failure to deliver a bounce message always caused it to be frozen, even if
2689 there was an errors_to setting on the router. The errors_to setting is now
2692 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2693 canonical form (fully expanded) before being placed in
2694 $sender_host_address.
2696 24. Updated eximstats to version 1.33
2698 25. Include certificate and key file names in error message when GnuTLS fails
2699 to set them up, because the GnuTLS error message doesn't include the name
2700 of the failing file when there is a problem reading it.
2702 26. Expand error message when OpenSSL has problems setting up cert/key files.
2705 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2706 (this can affect the format of dates).
2708 28. exim_tidydb, when checking for the continued existence of a message for
2709 which it has found a message-specific retry record, was not finding
2710 messages that were in split spool directories. Consequently, it was
2711 deleting retry records that should have stayed in existence.
2713 29. eximstats updated to version 1.35
2714 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2715 1.35 - bugfix such that pie charts by volume are generated correctly
2717 30. The SPA authentication driver was not abandoning authentication and moving
2718 on to the next authenticator when an expansion was forced to fail,
2719 contradicting the general specification for all authenticators. Instead it
2720 was generating a temporary error. It now behaves as specified.
2722 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2723 (the order specifies the preference for clients). The order is now AES256,
2724 AES128, 3DES, ARCFOUR128.
2726 31. Small patch to Sieve code - explicitly set From: when generating an
2729 32. Exim crashed if a remote delivery caused a very long error message to be
2730 recorded - for instance if somebody sent an entire SpamAssassin report back
2731 as a large number of 550 error lines. This bug was coincidentally fixed by
2732 increasing the size of one of Exim's internal buffers (big_buffer) that
2733 happened as part of the Exiscan merge. However, to be on the safe side, I
2734 have made the code more robust (and fixed the comments that describe what
2737 33. Some experimental protocols are using DNS PTR records for new purposes. The
2738 keys for these records are domain names, not reversed IP addresses. The
2739 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2740 leaves it alone. Component reversal etc. now happens only for IP addresses.
2743 34. The host_aton() function is supposed to be passed a string that is known
2744 to be a valid IP address. However, in the case of IPv6 addresses, it was
2745 not checking this. This is a hostage to fortune. Exim now panics and dies
2746 if the condition is not met. A case was found where this could be provoked
2747 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2748 components; fortuitously, this particular loophole had already been fixed
2749 by change 4.50/55 or 4.44/33 above.
2751 If there are any other similar loopholes, the new check in host_aton()
2752 itself should stop them being exploited. The report I received stated that
2753 data on the command line could provoke the exploit when Exim was running as
2754 exim, but did not say which command line option was involved. All I could
2755 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2756 running as the user.
2759 35. There was a buffer overflow vulnerability in the SPA authentication code
2760 (which came originally from the Samba project). I have added a test to the
2761 spa_base64_to_bits() function which I hope fixes it.
2764 36. The daemon start-up calls getloadavg() while still root for those OS that
2765 need the first call to be done as root, but it missed one case: when
2766 deliver_queue_load_max is set with deliver_drop_privilege. This is
2767 necessary for the benefit of the queue runner, because there is no re-exec
2768 when deliver_drop_privilege is set.
2770 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2771 was in use, was not putting the data itself into the right store pool;
2772 consequently, it could be overwritten for a subsequent message in the same
2773 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2776 38. Sometimes the final signoff response after QUIT could fail to get
2777 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2778 < 0 before doing a fflush(). This bug looks as though it goes back to the
2779 introduction of TLS in release 3.20, but "sometimes" must have been rare
2780 because the tests only now provoked it.
2786 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2787 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2788 Now it returns an integer. A number of calls were still expecting a T/F
2789 return. Fortuitously, in all cases, the tests worked in OK situations,
2790 which is the norm. However, things would have gone wrong on any write
2791 failures on the smtp file descriptor. This function is used when sending
2792 messages over SMTP and also when doing verify callouts.
2794 2. When Exim is called to do synchronous delivery of a locally submitted
2795 message (the -odf or -odi options), it no longer closes stderr before doing
2798 3. Implemented the mua_wrapper option.
2800 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2802 5. Implemented the functions header_remove(), header_testname(),
2803 header_add_at_position(), and receive_remove_recipient(), and exported them
2806 6. If an ACL "warn" statement specified the addition of headers, Exim already
2807 inserted X-ACL-Warn: at the start if there was no header name. However, it
2808 was not making this test for the second and subsequent header lines if
2809 there were newlines in the string. This meant that an invalid header could
2810 be inserted if Exim was badly configured.
2812 7. Allow an ACL "warn" statement to add header lines at the start or after all
2813 the Received: headers, as well as at the end.
2815 8. Added the rcpt_4xx retry error code.
2817 9. Added postmaster_mailfrom=xxx to callout verification option.
2819 10. Added mailfrom=xxxx to the callout verification option, for verify=
2822 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2823 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2825 12. Inserted some casts to stop certain compilers warning when using pointer
2826 differences as field lengths or precisions in printf-type calls (mostly
2827 affecting debugging statements).
2829 13. Added optional readline() support for -be (dynamically loaded).
2831 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2832 same clock tick as a message's arrival, so that its received time was the
2833 same as the "first fail" time on the retry record, and that message
2834 remained on the queue past the ultimate address timeout, every queue runner
2835 would try a delivery (because it was past the ultimate address timeout) but
2836 after another failure, the ultimate address timeout, which should have then
2837 bounced the address, did not kick in. This was a "< instead of <=" error;
2838 in most cases the first failure would have been in the next clock tick
2839 after the received time, and all would be well.
2841 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2842 being recognized when the domain list was tested by the match_domain
2843 condition in an expansion string.
2845 16. Added the ${str2b64: operator.
2847 17. Exim was always calling setrlimit() to set a large limit for the number of
2848 processes, without checking whether the existing limit was already
2849 adequate. (It did check for the limit on file descriptors.) Furthermore,
2850 errors from getrlimit() and setrlimit() were being ignored. Now they are
2851 logged to the main and panic logs, but Exim does carry on, to try to do its
2852 job under whatever limits there are.
2854 18. Imported PCRE 5.0.
2856 19. Trivial typo in log message " temporarily refused connection" (the leading
2859 20. If the log selector return_path_on_delivery was set and an address was
2860 redirected to /dev/null, the delivery process crashed because it assumed
2861 that a return path would always be set for a "successful" delivery. In this
2862 case, the whole delivery is bypassed as an optimization, and therefore no
2865 21. Internal re-arrangement: the function for sending a challenge and reading
2866 a response while authentication was assuming a zero-terminated challenge
2867 string. It's now changed to take a pointer and a length, to allow for
2868 binary data in such strings.
2870 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2872 23. Exim was not respecting finduser_retries when seeking the login of the
2873 uid under which it was called; it was always trying 10 times. (The default
2874 setting of finduser_retries is zero.) Also, it was sleeping after the final
2875 failure, which is pointless.
2877 24. Implemented tls_on_connect_ports.
2879 25. Implemented acl_smtp_predata.
2881 26. If the domain in control=submission is set empty, Exim assumes that the
2882 authenticated id is a complete email address when it generates From: or
2883 Sender: header lines.
2885 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2886 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2887 chown and chgrp in /bin and hostname in /usr/bin.
2889 28. Exim was keeping the "process log" file open after each use, just as it
2890 does for the main log. This opens the possibility of it remaining open for
2891 long periods when the USR1 signal hits a daemon. Occasional processlog
2892 errors were reported, that could have been caused by this. Anyway, it seems
2893 much more sensible not to leave this file open at all, so that is what now
2896 29. The long-running daemon process does not normally write to the log once it
2897 has entered its main loop, and it closes the log before doing so. This is
2898 so that log files can straightforwardly be renamed and moved. However,
2899 there are a couple of unusual error situations where the daemon does write
2900 log entries, and I had neglected to close the log afterwards.
2902 30. The text of an SMTP error response that was received during a remote
2903 delivery was being truncated at 512 bytes. This is too short for some of
2904 the long messages that one sometimes sees. I've increased the limit to
2907 31. It is now possible to make retry rules that apply only when a message has a
2908 specific sender, in particular, an empty sender.
2910 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2911 it possible to be selective about when SMTP synchronization is enforced.
2913 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2915 32. Implemented hosts_connection_nolog.
2917 33. Added an ACL for QUIT.
2919 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2922 35. Added mailbox_size and mailbox_filecount to appendfile.
2924 36. Added control = no_multiline_responses to ACLs.
2926 37. There was a bug in the logic of the code that waits for the clock to tick
2927 in the case where the clock went backwards by a substantial amount such
2928 that the microsecond fraction of "now" was more than the microsecond
2929 fraction of "then" (but the whole seconds number was less).
2931 38. Added support for the libradius Radius client library this is found on
2932 FreeBSD (previously only the radiusclient library was supported).
2938 1. When certain lookups returned multiple values in the form name=value, the
2939 quoting of the values was not always being done properly. Specifically:
2940 (a) If the value started with a double quote, but contained no whitespace,
2942 (b) If the value contained whitespace other than a space character (i.e.
2943 tabs or newlines or carriage returns) it was not quoted.
2944 This fix has been applied to the mysql and pgsql lookups by writing a
2945 separate quoting function and calling it from the lookup code. The fix
2946 should probably also be applied to nisplus, ibase and oracle lookups, but
2947 since I cannot test any of those, I have not disturbed their existing code.
2949 2. A hit in the callout cache for a specific address caused a log line with no
2950 reason for rejecting RCPT. Now it says "Previous (cached) callout
2951 verification failure".
2953 3. There was an off-by-one bug in the queryprogram router. An over-long
2954 return line was truncated at 256 instead of 255 characters, thereby
2955 overflowing its buffer with the terminating zero. As well as fixing this, I
2956 have increased the buffer size to 1024 (and made a note to document this).
2958 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2959 when Exim is waiting for an SMTP response from a remote server, Exim
2960 restarts its select() call on the socket, thereby resetting its timeout.
2961 This is not a problem when such interrupts are rare. Somebody set up a cron
2962 job to run exiwhat every 2 minutes, which is less than the normal select()
2963 timeout (5 or 10 minutes). This meant that the select() timeout never
2964 kicked in because it was always reset. I have fixed this by comparing the
2965 time when an interrupt arrives with the time at the start of the first call
2966 to select(). If more time than the timeout has elapsed, the interrupt is
2967 treated as a timeout.
2969 5. Some internal re-factoring in preparation for the addition of Sieve
2970 extensions (by MH). In particular, the "personal" test is moved to a
2971 separate function, and given an option for scanning Cc: and Bcc: (which is
2972 not set for Exim filters).
2974 6. When Exim created an email address using the login of the caller as the
2975 local part (e.g. when creating a From: or Sender: header line), it was not
2976 quoting the local part when it contained special characters such as @.
2978 7. Installed new OpenBSD configuration files.
2980 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2981 try to make them clearer.
2983 9. Callout options, other than the timeout value, were being ignored when
2984 verifying sender addresses in header lines. For example, when using
2986 verify = header_sender/callout=no_cache
2988 the cache was (incorrectly) being used.
2990 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2991 only the Cygwin environment.
2993 11. When return_path_on_delivery was set as a log selector, if different remote
2994 addresses in the same message used different return paths and parallel
2995 remote delivery occurred, the wrong values would sometimes be logged.
2996 (Whenever a remote delivery process finished, the return path value from
2997 the most recently started remote delivery process was logged.)
2999 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3000 lines when AUTH and/or TLS are in use. This is the "received protocol"
3001 field. Exim used to use "asmtp" for authenticated SMTP, without any
3002 indication (in the protocol name) for TLS use. Now it follows the RFC and
3003 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3004 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3005 names appear in log lines as well as in Received: header lines.
3007 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3008 extensions, and comparison tests, and to fix some bugs.
3010 14. Changes to the "personal" filter test:
3012 (1) The test was buggy in that it was just doing the equivalent of
3013 "contains" tests on header lines. For example, if a user's address was
3014 anne@some.where, the "personal" test would incorrectly be true for
3016 To: susanne@some.where
3018 This test is now done by extracting each address from the header in turn,
3019 and checking the entire address. Other tests that are part of "personal"
3020 are now done using regular expressions (for example, to check local parts
3021 of addresses in From: header lines).
3023 (2) The list of non-personal local parts in From: addresses has been
3024 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3025 taken from the Sieve specification recommendations.
3027 (3) If the message contains any header line starting with "List-" it is
3028 treated as non-personal.
3030 (4) The test for "circular" in the Subject: header line has been removed
3031 because it now seems ill-conceived.
3033 15. Minor typos in src/EDITME comments corrected.
3035 16. Installed latest exipick from John Jetmore.
3037 17. If headers_add on a router specified a text string that was too long for
3038 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3039 of string_sprintf() is now avoided.
3041 18. $message_body_size was not set (it was always zero) when running the DATA
3042 ACL and the local_scan() function.
3044 19. For the "mail" command in an Exim filter, no default was being set for
3045 the once_repeat time, causing a random time value to be used if "once" was
3046 specified. (If the value happened to be <= 0, no repeat happened.) The
3047 default is now 0s, meaning "never repeat". The "vacation" command was OK
3048 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3049 (I found it when inspecting the code).
3051 20. There is now an overall timeout for performing a callout verification. It
3052 defaults to 4 times the callout timeout, which applies to individual SMTP
3053 commands during the callout. The overall timeout applies when there is more
3054 than one host that can be tried. The timeout is checked before trying the
3055 next host. This prevents very long delays if there are a large number of
3056 hosts and all are timing out (e.g. when the network connections are timing
3057 out). The value of the overall timeout can be changed by specifying an
3058 additional sub-option for "callout", called "maxwait". For example:
3060 verify = sender/callout=5s,maxwait=20s
3062 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3063 to the end before writing, but this should make it even safer).
3065 22. Exim was forgetting that it had advertised PIPELINING for the second and
3066 subsequent messages on an SMTP connection. It was also not resetting its
3067 memory on STARTTLS and an internal HELO.
3069 23. When Exim logs an SMTP synchronization error within a session, it now
3070 records whether PIPELINING has been advertised or not.
3072 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3073 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3074 rather than long int.
3076 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3078 26. Added the never_mail option to autoreply.
3084 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3085 crash if the getsockname() call failed; this can happen if a connection is
3086 closed very soon after it is established. The problem was simply in the
3087 order in which certain operations were done, causing Exim to try to write
3088 to the SMTP stream before it had set up the file descriptor. The bug has
3089 been fixed by making things happen in the correct order.
3095 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3096 before the connection was closed, thus losing the rejection response.
3098 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3099 some early Solaris releases, but causes trouble in current releases where
3100 socklen_t is defined.
3102 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3105 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3108 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3109 that releasing the top of it at the end releases what was used for sub-
3110 expansions (unless the block got too big). However, discard this block if
3111 the first thing is a variable or header, so that we can use its block when
3112 it is dynamic (useful for very large $message_headers, for example).
3114 6. Lookups now cache *every* query, not just the most recent. A new, separate
3115 store pool is used for this. It can be recovered when all lookup caches are
3116 flushed. Lookups now release memory at the end of their result strings.
3117 This has involved some general refactoring of the lookup sources.
3119 7. Some code has been added to the store_xxx() functions to reduce the amount
3120 of flapping under certain conditions.
3122 8. log_incoming_interface used to affect only the <= reception log lines. Now
3123 it causes the local interface and port to be added to several more SMTP log
3124 lines, for example "SMTP connection from", and rejection lines.
3126 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3128 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3130 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3131 could be overwritten at the end of the current message (or the start of a
3132 new message if it was set in a HELO ACL). The value is now preserved for
3133 the duration of the SMTP connection.
3135 12. If a transport had a headers_rewrite setting, and a matching header line
3136 contained an unqualified address, that address was qualified, even if it
3137 did not match any rewriting rules. The underlying bug was that the values
3138 of the flags that permit the existence of unqualified sender and recipient
3139 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3140 non-local messages, and by -bnq for local messages) were not being
3141 preserved with the message after it was received.
3143 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3144 "next input=" as part of the text comprising the host identity instead of
3145 the correct text. The code was using the same buffer for two different
3146 strings. However, depending on which order the printing function evaluated
3147 its arguments, the bug did not always show up. Under Linux, for example, my
3148 test suite worked just fine.
3150 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3151 doesn't work with some older versions of Perl. It has been changed to "my",
3152 which in any case is probably the better facility to use.
3154 15. A really picky compiler found some instances of statements for creating
3155 error messages that either had too many or two few arguments for the format
3158 16. The size of the buffer for calls to the DNS resolver has been increased
3159 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3160 for addresses that have a lot of PTR records. This alleviates a problem; it
3161 does not fully solve it.
3163 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3164 buffer now truncates the list and logs the incident, which is the same
3165 action as happens when Exim is looking up a host name and its aliases.
3166 Previously in this situation something unpredictable would happen;
3167 sometimes it was "internal error: store_reset failed".
3169 18. If a server dropped the connection unexpectedly when an Exim client was
3170 using GnuTLS and trying to read a response, the client delivery process
3171 crashed while trying to generate an error log message.
3173 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3174 string, for example:
3176 warn message = H1: something\nH2: something
3178 the text was added as a single header line from Exim's point of view
3179 though it ended up OK in the delivered message. However, searching for the
3180 second and subsequent header lines using $h_h2: did not work. This has been
3181 fixed. Similarly, if a system filter added multiple headers in this way,
3182 the routers could not see them.
3184 20. Expanded the error message when iplsearch is called with an invalid key to
3185 suggest using net-iplsearch in a host list.
3187 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3188 ACLs are no longer actually imposed (and a message to that effect is
3191 22. If a "gecos" field in a passwd entry contained escaped characters, in
3192 particular, if it contained a \" sequence, Exim got it wrong when building
3193 a From: or a Sender: header from that name. A second bug also caused
3194 incorrect handling when an unquoted " was present following a character
3195 that needed quoting.
3197 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3198 was not being matched caselessly.
3200 24. Arranged for all hyphens in the exim.8 source to be escaped with
3203 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3204 properly. Recipient callout cache records were still being keyed to include
3205 the sender, even when use_sender was set false. This led to far more
3206 callouts that were necessary. The sender is no longer included in the key
3207 when use_sender is false.
3209 26. Added "control = submission" modifier to ACLs.
3211 27. Added the ${base62d: operator to decode base 62 numbers.
3213 28. dnsdb lookups can now access SRV records.
3215 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3216 the configuration file.
3218 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3219 (-v) mode. This makes the output for a verbose queue run more intelligible.
3221 31. Added a use_postmaster feature to recipient callouts.
3223 32. Added the $body_zerocount variable, containing the number of binary zero
3224 bytes in the message body.
3226 33. The time of last modification of the "new" subdirectory is now used as the
3227 "mailbox time last read" when there is a quota error for a maildir
3230 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3232 35. Added +ignore_unknown as a special item in host lists.
3234 36. Code for decoding IPv6 addresses in host lists is now included, even if
3235 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3236 address was recognized as an IP address, but was then not correctly decoded
3237 into binary, causing unexpected and incorrect effects when compared with
3244 1. Very minor rewording of debugging text in manualroute to say "list of
3245 hosts" instead of "hostlist".
3247 2. If verify=header_syntax was set, and a header line with an unqualified
3248 address (no domain) and a large number of spaces between the end of the
3249 name and the colon was received, the reception process suffered a buffer
3250 overflow, and (when I tested it) crashed. This was caused by some obsolete
3251 code that should have been removed. The fix is to remove it!
3253 3. When running in the test harness, delay a bit after writing a bounce
3254 message to get a bit more predictability in the log output.
3256 4. Added a call to search_tidyup() just before forking a reception process. In
3257 theory, someone could use a lookup in the expansion of smtp_accept_max_
3258 per_host which, without the tidyup, could leave open a database connection.
3260 5. Added the variables $recipient_data and $sender_data which get set from a
3261 lookup success in an ACL "recipients" or "senders" condition, or a router
3262 "senders" option, similar to $domain_data and $local_part_data.
3264 6. Moved the writing of debug_print from before to after the "senders" test
3267 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3268 problems for message scanning, either using a data ACL, or using
3269 local_scan() because the Received: header was not generated till after they
3270 were called (in order to set the time as the time of reception completion).
3271 I have revised the way this works. The header is now generated after the
3272 body is received, but before the ACL or local_scan() are called. After they
3273 are run, the timestamp in the header is updated.
3279 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3280 before starting a queue runner without re-exec. This happened only when
3281 deliver_drop_privilege was set or when the Exim user was set to root. The
3282 effect of the bug was that timeouts during subsequent deliveries caused
3283 crashes instead of being properly handled. The handler is now left at its
3284 default (and expected) setting.
3286 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3287 message, again when deliver_drop_privilege is set or Exim is run as root.
3288 The bug described in (1) was not present in this case, but the tidying up
3289 of the other signals was missing. I have made the two cases consistent.
3291 3. The ignore_target_hosts setting on a manualroute router was being ignored
3292 for hosts that were looked up using the /MX notation.
3294 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3297 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3298 operated on the sender address. After changing the $sender_address to <>
3299 for the sender address verify, Exim was re-instated it as the original
3300 (before rewriting) address, but remembering that it had rewritten it, so it
3301 wasn't rewriting it again. This bug also had the effect of breaking the
3302 sender address verification caching when the sender address was rewritten.
3304 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3305 This has been changed so that if the ip literal address matches
3306 ignore_target_hosts, the router declines.
3308 7. Added expansion conditions match_domain, match_address, and match_local_
3309 part (NOT match_host).
3311 8. The placeholder for the Received: header didn't have a length field set.
3313 9. Added code to Exim itself and to exim_lock to test for a specific race
3314 condition that could lead to file corruption when using MBX delivery. The
3315 issue is with the lockfile that is created in /tmp. If this file is removed
3316 after a process has opened it but before that process has acquired a lock,
3317 there is the potential for a second process to recreate the file and also
3318 acquire a lock. This could lead to two Exim processes writing to the file
3319 at the same time. The added code performs the same test as UW imapd; it
3320 checks after acquiring the lock that its file descriptor still refers to
3321 the same named file.
3323 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3324 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3327 11. Added the smtp_active_hostname option. If used, this will typically be made
3328 to depend on the incoming interface address. Because $interface_address is
3329 not set up until the daemon has forked a reception process, error responses
3330 that can happen earlier (such as "too many connections") no longer contain
3333 12. If an expansion in a condition on a "warn" statement fails because a lookup
3334 defers, the "warn" statement is abandoned, and the next ACL statement is
3335 processed. Previously this caused the whole ACL to be aborted.
3337 13. Added the iplsearch lookup type.
3339 14. Added ident_timeout as a log selector.
3341 15. Added tls_certificate_verified as a log selector.
3343 16. Added a global option tls_require_ciphers (compare the smtp transport
3344 option of the same name). This controls incoming TLS connections.
3346 17. I finally figured out how to make tls_require_ciphers do a similar thing
3347 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3348 before starting the TLS session.
3350 18. Tabs are now shown as \t in -bP output.
3352 19. If the log selector return_path_on_delivery was set, Exim crashed when
3353 bouncing a message because it had too many Received: header lines.
3355 20. If two routers both had headers_remove settings, and the first one included
3356 a superfluous trailing colon, the final name in the first list and the
3357 first name in the second list were incorrectly joined into one item (with a
3358 colon in the middle).
3364 1. Added -C and -D options to the exinext utility, mainly to make it easier
3365 to include in the automated testing, but these could be helpful when
3366 multiple configurations are in use.
3368 2. The exinext utility was not formatting the output nicely when there was
3369 an alternate port involved in the retry record key, nor when there was a
3370 message id as well (for retries that were specific to a specific message
3371 and a specific host). It was also confused by IPv6 addresses, because of
3372 the additional colons they contain. I have fixed the IPv4 problem, and
3373 patched it up to do a reasonable job for IPv6.
3375 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3376 delivery, the log line now contains "pipelined" if PIPELINING was used.
3378 4. An SMTP transport process used to panic and die if the bind() call to set
3379 an explicit outgoing interface failed. This has been changed; it is now
3380 treated in the same way as a connect() failure.
3382 5. A reference to $sender_host_name in the part of a conditional expansion
3383 that was being skipped was still causing a DNS lookup. This no longer
3386 6. The def: expansion condition was not recognizing references to header lines
3387 that used bh_ and bheader_.
3389 7. Added the _cache feature to named lists.
3391 8. The code for checking quota_filecount in the appendfile transport was
3392 allowing one more file than it should have been.
3394 9. For compatibility with Sendmail, the command line option
3402 and sets the incoming protocol and host name (for trusted callers). The
3403 host name and its colon can be omitted when only the protocol is to be set.
3404 Note the Exim already has two private options, -pd and -ps, that refer to
3405 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3406 "s", but I don't think that's a major issue.
3408 10. A number of refactoring changes to the code, none of which should affect
3411 (a) The number of logging options was getting close to filling up the
3412 32-bit word that was used as a bit map. I have split them into two classes:
3413 those that are passed in the argument to log_write(), and those that are
3414 only ever tested independently outside of that function. These are now in
3415 separate 32-bit words, so there is plenty of room for expansion again.
3416 There is no change in the user interface or the logging behaviour.
3418 (b) When building, for example, log lines, the code previously used a
3419 macro that called string_cat() twice, in order to add two strings. This is
3420 not really sufficiently general. Furthermore, there was one instance where
3421 it was actually wrong because one of the argument was used twice, and in
3422 one call a function was used. (As it happened, calling the function twice
3423 did not affect the overall behaviour.) The macro has been replaced by a
3424 function that can join an arbitrary number of extra strings onto a growing
3427 (c) The code for expansion conditions now uses a table and a binary chop
3428 instead of a serial search (which was left over from when there were very
3429 few conditions). Also, it now recognizes conditions like "pam" even when
3430 the relevant support is not compiled in: a suitably worded error message is
3431 given if an attempt is made to use such a condition.
3433 11. Added ${time_interval:xxxxx}.
3435 12. A bug was causing one of the ddress fields not to be passed back correctly
3436 from remote delivery subprocesses. The field in question was not being
3437 subsequently used, so this caused to problems in practice.
3439 13. Added new log selectors queue_time and deliver_time.
3441 14. Might have fixed a bug in maildirsizefile handling that threw up
3442 "unexpected character" debug warnings, and recalculated the data
3443 unnecessarily. In any case, I expanded the warning message to give more
3446 15. Added the message "Restricted characters in address" to the statements in
3447 the default ACL that block characters like @ and % in local parts.
3449 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3450 Three changes have been made:
3452 (a) There was a serious bug; a negative response to MAIL caused the whole
3453 recipient domain to be cached as invalid, thereby blocking all messages
3454 to all local parts at the same domain, from all senders. This bug has
3455 been fixed. The domain is no longer cached after a negative response to
3456 MAIL if the sender used is not empty.
3458 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3461 (c) A new callout option, "use_sender" has been added for people who want
3462 the modified behaviour.
3468 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3471 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3472 indeed breaks things for older releases.
3474 3. Added additional logging to the case where there is a problem reading data
3475 from a filter that is running in a subprocess using a pipe, in order to
3476 try to track down a specific problem.
3478 4. Testing facility fudge: when running in the test harness and attempting
3479 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3480 getting "No route to host". Convert this to a timeout.
3482 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3485 6. Some OS don't have socklen_t but use size_t instead. This affects the
3486 fifth argument of getsockopt() amongst other things. This is now
3487 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3488 can be set for individual OS. I have set it for SunOS5, OSF1, and
3489 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3490 some earlier ones do not.
3492 7. Change 4.30/15 was not doing the test caselessly.
3494 8. The standard form for an IPv6 address literal was being rejected by address
3495 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3496 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3497 this, as well as the form without the "IPv6" on the front (but only when
3498 address literals are enabled, of course).
3500 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3502 10. Exim crashed if a message with an empty sender address specified by -f
3503 encountered a router with an errors_to setting. This could be provoked only
3504 by a command such as
3508 where an empty string was supplied; "<>" did not hit this bug.
3510 11. Installed PCRE release 4.5.
3512 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3513 remained set. It is now erased.
3515 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3516 times from message ids (which are base 36 rather than the normal 62).
3518 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3519 were being counted as actual protocol errors, and logged if the log
3520 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3521 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3522 and DATA following a set of rejected RCPTs do not count as protocol errors.
3523 In other words, Exim assumes they were pipelined, though this may not
3524 actually be the case. Of course, in all cases the client gets an
3525 appropriate error code.
3527 15. If a lookup fails in an ACL condition, a message about the failure may
3528 be available; it is used if testing the ACL cannot continue, because most
3529 such messages specify what the cause of the deferral is. However, some
3530 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3531 that caused an old message to be retained and used if a later statement
3532 caused a defer, replacing the real cause of the deferral.
3534 16. If an IP address had so many PTR records that the DNS lookup buffer
3535 was not large enough to hold them, Exim could crash while trying to process
3536 the truncated data. It now detects and logs this case.
3538 17. Further to 4.21/58, another change has been made: if (and only if) the
3539 first line of a message (the first header line) ends with CRLF, a bare LF
3540 in a subsequent header line has a space inserted after it, so as not to
3541 terminate the header.
3543 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3544 unnecessarily, used atoi() instead of strtol(), and didn't check the
3545 termination when getting file sizes from file names by regex.
3547 19. Completely re-implemented the support for maildirsize files, in the light
3548 of a number of problems with the previous contributed implementation
3549 (4.30/29). In particular:
3551 . If the quota is zero, the maildirsize file is maintained, but no quota is
3554 . If the maildir directory does not exist, it is created before any attempt
3555 to write a maildirsize file.
3557 . The quota value in the file is just a cache; if the quota is changed in
3558 the transport, the new value overrides.
3560 . A regular expression is available for excluding directories from the
3563 20. The autoreply transport checks the characters in options that define the
3564 message's headers; it allows continued headers, but it was checking with
3565 isspace() after an embedded newline instead of explicitly looking for a
3568 21. If all the "regular" hosts to which an address was routed had passed their
3569 expiry times, and had not reached their retry times, the address was
3570 bounced, even if fallback hosts were defined. Now Exim should go on to try
3573 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3574 equivalent code in the SMTP transport. Some hosts send humungous responses
3575 to HELO/EHLO, more than 1024 it seems.
3577 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3578 gives compiler warnings in some environments. I've now done it "properly",
3581 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3582 (because of problems with the built-in one) was declared to return uschar *
3583 instead of char *, causing compiler failure.
3585 25. Fixed a file descriptor leak when processing alias/forward files.
3587 26. Fixed a minor format string issue in dbfn.c.
3589 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3591 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3592 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3595 29. When a sender address is verified, it is cached, to save repeating the test
3596 when there is more than one recipient in a message. However, when the
3597 verification involves a callout, it is possible for different callout
3598 options to be set for different recipients. It is too complicated to keep
3599 track of this in the cache, so now Exim always runs a verification when a
3600 callout is required, relying on the callout cache for the optimization.
3601 The overhead is duplication of the address routing, but this should not be
3604 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3605 to be verified with callout=postmaster, and the main callout worked but the
3606 postmaster check failed, the verification correctly failed. However, if a
3607 subsequent RCPT command asked for sender verification *without* the
3608 postmaster check, incorrect caching caused this verification also to fail,
3611 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3612 it was not caching the DNS options (qualify_single, search_parents) that
3613 were used when the lookup failed. A subsequent lookup with different
3614 options therefore always gave the same answer, though there were cases
3615 where it should not have. (Example: a "domains = !$mx_any" option on a
3616 dnslookup router: the "domains" option is always processed without any
3617 widening, but the router might have qualify_single set.) Now Exim uses the
3618 cached value only when the same options are set.
3620 32. Added John Jetmore's "exipick" utility to the distribution.
3622 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3623 than a timeout (e.g. a certificate is required, and is not provided), an
3624 Exim server now closes the connection immediately. Previously it waited for
3625 the client to close - but if the client is SSL, it seems that they each
3626 wait for each other, leading to a delay before one of them times out.
3628 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3629 maintained 0.8.x compatibility because I don't think many are using it, and
3630 it is clearly obsolete.
3632 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3635 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3636 client certificate was expired. A simple patch fixes this, though I don't
3637 understand the full logic of why the verify callback is called multiple
3640 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3641 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3642 which causes problems with some clients (such as the Certicom SSL Plus
3643 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3644 disables the coutermeasure allowing Eudora to connect."
3646 38. Exim was not checking that a write() to a log file succeeded. This could
3647 lead to Bad Things if a log got too big, in particular if it hit a file
3648 size limit. Exim now panics and dies if it cannot write to a log file, just
3649 as it does if it cannot open a log file.
3651 39. Modified OS/Makefile-Linux so that it now contains
3653 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3655 The two -D definitions ensure that Exim is compiled with large file
3656 support, which makes it possible to handle log files that are bigger than
3659 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3660 instance) a domain was checked against a named list that involved a lookup,
3661 causing $domain_data to be set, then another domain was checked against the
3662 same list, then the first domain was re-checked, the value of $domain_data
3663 after the final check could be wrong. In particular, if the second check
3664 failed, it could be set empty. This bug probably also applied to
3667 41. The strip_trailing_dot option was not being applied to the address given
3668 with the -f command-line option.
3670 42. The code for reading a message's header from the spool was incrementing
3671 $received_count, but never initializing it. This meant that the value was
3672 incorrect (doubled) while delivering a message in the same process in which
3673 it was received. In the most common configuration of Exim, this never
3674 happens - a fresh exec is done - but it can happen when
3675 deliver_drop_privilege is set.
3677 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3678 it now includes up to 150 characters of the unexpected data in the log
3681 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3682 and building data strings. The size of both of these buffers was 10 000
3683 bytes - far larger than anybody would *ever* want, thought I. Needless to
3684 say, somebody hit the limit. I have increased the maximum line length to
3685 20 000 and the maximum data length of concatenated lines to 100 000. I have
3686 also fixed two bugs, because there was no checking on these buffers. Tsk,
3687 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3688 buffer is too small.
3690 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3691 lsearch lookups. Now it does.
3693 46. When parsing a route_list item in a manualroute router, a fixed-length
3694 buffer was used for the list of hosts. I made this 1024 bytes long,
3695 thinking that nobody would ever have a list of hosts that long. Wrong.
3696 Somebody had a whole pile of complicated expansion conditions, and the
3697 string was silently truncated, leading to an expansion error. It turns out
3698 that it is easier to change to an unlimited length (owing to other changes
3699 that have happened since this code was originally written) than to build
3700 structure for giving a limitation error. The length of the item that
3701 expands into the list of hosts is now unlimited.
3703 47. The lsearch lookup could not handle data where the length of text line was
3704 more than 4095 characters. Such lines were truncated, leading to shortened
3705 data being returned. It should now handle lines of any length.
3707 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3708 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3711 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3712 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3713 no longer happens in comment lines. A list of replacements is now placed
3714 at the head of all of the source files, except those whose only change is
3715 to replace PERL_COMMAND in the very first #! line.
3717 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3718 messages on the queue, with a bottom-up merge sort, using code contributed
3719 by Michael Haardt. This should make operations like -bp somewhat faster on
3720 large queues. It won't affect queue runners, except when queue_run_in_order
3723 51. Installed eximstats 1.31 in the distribution.
3725 52. Added support for SRV lookups to the dnslookup router.
3727 53. If an ACL referred to $message_body or $message_body_end, the value was not
3728 reset for any messages that followed in the same SMTP session.
3730 54. The store-handling optimization for building very long strings was not
3731 differentiating between the different store pools. I don't think this
3732 actually made any difference in practice, but I've tidied it.
3734 55. While running the routers to verify a sender address, $sender_address
3735 was still set to the sender address. This is wrong, because when routing to
3736 send a bounce to the sender, it would be empty. Therefore, I have changed
3737 it so that, while verifying a sender address, $sender_address is set to <>.
3738 (There is no change to what happens when verifying a recipient address.)
3740 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3741 target A or AAAA records (if not already returned) without resetting the
3742 qualify_single or search_parents options of the DNS resolver. These are
3743 inappropriate in this case because the targets of MX and SRV records must
3744 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3745 target that, when qualified, matched something in the local domain. These
3746 two options are now turned off when doing these lookups.
3748 57. It seems that at least some releases of Reiserfs (which does not have the
3749 concept of a fixed number of inodes) returns zero and not -1 for the
3750 number of available inodes. This interacted badly with check_spool_inodes,
3751 which assumed that -1 was the "no such thing" setting. What I have done is
3752 to check that the total number of inodes is greater than zero before doing
3753 the test of how many are available.
3755 58. When a "warn" ACL statement has a log_message modifier, the message is
3756 remembered, and not repeated. This is to avoid a lot of repetition when a
3757 message has many recipients that cause the same warning to be written.
3758 Howewer, Exim was preserving the list of already written lines for an
3759 entire SMTP session, which doesn't seem right. The memory is now reset if a
3760 new message is started.
3762 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3763 debugging output unless log_rewrite was also set.
3765 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3766 of (int)(handle) when we know that handle contains (void *)(-1).
3768 61. The Exim daemon panic-logs an error return when it closes the incoming
3769 connection. However "connection reset by peer" seems to be common, and
3770 isn't really an error worthy of noting specially, so that particular error
3773 62. When Exim is trying to find all the local interfaces, it used to panic and
3774 die if the ioctl to get the interface flags failed. However, it seems that
3775 on at least one OS (Solaris 9) it is possible to have an interface that is
3776 included in the list of interfaces, but for which you get a failure error
3777 for this call. This happens when the interface is not "plumbed" into a
3778 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3779 failure of the "get flags" call assumes that the interface is down.
3781 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3782 makes life easier for people who are doing arithmetic on fields extracted
3783 from dates, where you often get leading zeros that should not be
3784 interpreted as octal.
3786 64. Added qualify_domain to the redirect router, to override the global
3789 65. If a pathologically long header line contained very many addresses (the
3790 report of this problem mentioned 10 000) and each of them was rewritten,
3791 Exim could use up a very large amount of memory. (It kept on making new
3792 copies of the header line as it rewrote, and never released the old ones.)
3793 At the expense of a bit more processing, the header rewriting function has
3794 been changed so that it no longer eats memory in this way.
3796 66. The generation of the Received: header has been moved from the time that a
3797 message starts to be received, to the time that it finishes. The timestamp
3798 in the Received: header should now be very close to that of the <= log
3799 line. There are two side-effects of this change:
3801 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3802 logged header lines no longer include the local Received: line, because
3803 it has not yet been created. The same applies to a copy of the message
3804 that is returned to a non-SMTP sender when a message is rejected.
3806 (b) When a filter file is tested using -bf, no additional Received: header
3807 is added to the test message. After some thought, I decided that this
3810 This change does not affect the value of $received_for. It is still set
3811 after address rewriting, but before local_scan() is called.
3813 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3815 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3816 gave an unhelpful panic error message, and a defer error. I have managed to
3817 change this behaviour so that it now rejects any supplied certificate,
3818 which seems right, as the list of acceptable certificates is empty.
3820 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3821 gave an unhelpful defer error. I have not managed to make this reject any
3822 supplied certificates, but the error message it gives is "no certificate
3823 supplied", which is not helpful.
3825 70. exigrep's output now also includes lines that are not associated with any
3826 message, but which match the given pattern. Implemented by a patch from
3827 Martin Sluka, which also tidied up the Perl a bit.
3829 71. Recipient callout verification, like sender verification, was using <> in
3830 the MAIL FROM command. This isn't really the right thing, since the actual
3831 sender may affect whether the remote host accepts the recipient or not. I
3832 have changed it to use the actual sender in the callout; this means that
3833 the cache record is now keyed on a recipient/sender pair, not just the
3834 recipient address. There doesn't seem to be a real danger of callout loops,
3835 since a callout by the remote host to check the sender would use <>.
3836 [SEE ABOVE: changed after hitting problems.]
3838 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3839 temporary errors. However, in the case of such a code being given after
3840 the end of a data transmission (i.e. after ".") Exim was failing to write
3841 a retry record for the message. (Yes, there was some broken host that was
3842 actually sending 8xx at this point.)
3844 73. An unknown lookup type in a host list could cause Exim to panic-die when
3845 the list was checked. (An example that provoked this was putting <; in the
3846 middle of a list instead of at the start.) If this happened during a DATA
3847 ACL check, a -D file could be left lying around. This kind of configuration
3848 error no longer causes Exim to die; instead it causes a defer errror. The
3849 incident is still logged to the main and panic logs.
3851 74. Buglet left over from Exim 3 conversion. The message "too many messages
3852 in one connection" was written to the rejectlog but not the mainlog, except
3853 when address rewriting (yes!) was being logged.
3855 75. Added write_rejectlog option.
3857 76. When a system filter was run not as root (that is, when system_filter_user
3858 was set), the values of the $n variables were not being returned to the
3859 main process; thus, they were not subsequently available in the $sn
3862 77. Added +return_path_on_delivery log selector.
3864 78. A connection timeout was being treated differently from recipients deferred
3865 when testing hosts_max_try with a message that was older than the host's
3866 retry timeout. (The host should not be counted, thus allowing all hosts to
3867 be tried at least once before bouncing.) This may have been the cause of an
3868 occasionally reported bug whereby a message would remain on the queue
3869 longer than the retry timeout, but would be bounced if a delivery was
3870 forced. I say "may" because I never totally pinned down the problem;
3871 setting up timeout/retry tests is difficult. See also the next item.
3873 79. The ultimate address timeout was not being applied to errors that involved
3874 a combination of host plus message (for example, a timeout on a MAIL
3875 command). When an address resolved to a number of possible hosts, and they
3876 were not all tried for each delivery (e.g. because of hosts_max_try), a
3877 message could remain on the queue longer than the retry timeout.
3879 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3882 81. Fixed an obscure SMTP outgoing bug which required at least the following
3883 conditions: (a) there was another message waiting for the same server;
3884 (b) the server returned 5xx to all RCPT commands in the first message so
3885 that the message was not completed; (c) the server dropped the connection
3886 or gave a negative response to the RSET that Exim sends to abort the
3887 transaction. The observed case was a dropped connection after DATA that had
3888 been sent in pipelining mode. That is, the server had advertised PIPELINING
3889 but was not implementing it correctly. The effect of the bug was incorrect
3890 behaviour, such as trying another host, and this could lead to a crash.
3896 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3897 and daemon.c were passed as pointers to ints; they should have been
3898 pointers to socklen_t variables (which are typically unsigned ints).
3900 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3903 3. Fixed a really odd bug that affected only the testing scheme; patching a
3904 certain fixed string in the binary changed the value of another string that
3905 happened to be identical to the end of the original first string.
3907 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3908 name", it returns that address as the IP address. On some operating
3909 systems (e.g. Solaris), it also passes back the IP address string as the
3910 "host name". However, on others (e.g. Linux), it passes back an empty
3911 string. Exim wasn't checking for this, and was changing the host name to an
3912 empty string, assuming it had been canonicized.
3914 5. Although rare, it is permitted to have more than one PTR record for a given
3915 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3916 all the names associated with an address, because they do in Solaris.
3917 However, it seems that they do not in Linux for data that comes from the
3918 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3919 I found this out when I moved to a new Linux workstation and tried to run
3920 the Exim test suite.
3922 To get round this problem I have changed the code so that it now does its
3923 own call to the DNS to look up PTR records when searching for a host name.
3924 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3925 addresses that are only in /etc/hosts are still found.
3927 This behaviour is, however, controlled by an option called host_lookup_
3928 order, which defaults to "bydns:byaddr". If people want to use the other
3929 order, or indeed, just use one or the other means of lookup, they can
3930 specify it in this variable.
3932 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3933 some operating systems, this comes back from gethostbyaddr() as an empty
3934 string, and this is what Exim used to test for. However, it seems that in
3935 other systems, "." is yielded. Exim now tests for this case too.
3937 7. The values of check_spool_space and check_log_space are now held internally
3938 as a number of kilobytes instead of an absolute number of bytes. If a
3939 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3940 kilobyte. This means that much larger values can be stored.
3942 8. Exim monitor: an attempt to get the action menu when not actually pointing
3943 at a message produces an empty menu entitled "No message selected". This
3944 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3945 no entries in it ("Shell widget menu has zero width and/or height"). So I
3946 have added a single, blank menu entry in this case.
3948 9. Added ${quote_local_part.
3950 10. MIME decoding is now applied to the contents of Subject: header lines when
3953 11. Now that a reference to $sender_host_address automatically causes a reverse
3954 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3955 host lookup before query-style lookups in lists that might use this
3956 variable. This has therefore been abolished, and the "net-" prefix is no
3957 longer necessary for query-style lookups.
3959 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3960 have been a typo for LFLAGS, so it has been changed.
3962 13. The install script calls Exim with "-C /dev/null" in order to find the
3963 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3964 to be output. Howeve, since Exim outputs its version number before the
3965 error, it didn't break the script. It just looked ugly. I fixed this by
3966 always allowing "-C /dev/null" if the caller is root.
3968 14. Ignore overlarge ACL variable number when reading spool file - insurance
3969 against a later release with more variables having written the file.
3971 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3972 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3973 this, as well as the form without the "IPv6" on the front.
3975 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3976 OS/Makefile-Darwin file.
3978 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3979 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3981 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3982 variable. However, some OS use ints for inodes. I've added cast to long int
3983 to get rid of the compiler warning.
3985 19. I had forgotten to lock out "/../" in configuration file names when
3986 ALT_CONFIG_PREFIX was set.
3988 20. Routers used for verification do not need to specify transports. However,
3989 if such a router generated a host list, and callout was configured, Exim
3990 crashed, because it could not find a port number from the (non-existent)
3991 transport. It now assumes port 25 in this circumstance.
3993 21. Added the -t option to exigrep.
3995 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3996 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3997 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3998 them is set, LOOKUP_LSEARCH is forced.
4000 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4002 24. Added sender and host information to the "rejected by local_scan()" log
4003 line; previously there was no indication of these.
4005 25. Added .include_if_exists.
4007 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4008 Linux. It turns out that not all file systems support this. Apparently some
4009 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4010 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4011 not supported on the file descriptor, is now ignored when Exim is trying to
4012 sync a directory. This applies only to Linux.
4014 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4016 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4017 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4020 29. Added support for maildirsize files from supplied patch (modified a bit).
4022 30. The use of :fail: followed by an empty string could lead Exim to respond to
4023 sender verification failures with (e.g.):
4025 550 Verification failed for <xxx>
4026 550 Sender verify failed
4028 where the first response line was missing the '-' that indicates it is not
4029 the final line of the response.
4031 31. The loop for finding the name of the user that called Exim had a hardwired
4032 limit of 10; it now uses the value of finduser_retries, which is used for
4033 all other user lookups.
4035 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4038 33. Exim was neglecting to zero errno before one call of strtol() when
4039 expanding a string and expecting an integer value. On some systems this
4040 resulted in spurious "integer overflow" errors. Also, it was casting the
4041 result into an int without checking.
4043 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4044 did not work. The code looks as if it has *never* worked, though it appears
4045 to have been documented since at least releast 1.62. I have made it work.
4047 35. The "timeout_DNS" error in retry rules, also documented since at least
4048 1.62, also never worked. As it isn't clear exactly what this means, and
4049 clearly it isn't a major issue, I have abolished the feature by treating it
4050 as "timeout", and writing a warning to the main and panic logs.
4052 36. The display of retry rules for -brt wasn't always showing the error code
4055 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4056 timeout_connect_A, timeout_connect_MX.
4058 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4059 to the empty sender.
4061 39. The daemon was not analysing the content of -oX till after it had closed
4062 stderr and disconnected from the controlling terminal. This meant that any
4063 syntax errors were only noted on the panic log, and the return code from
4064 the command was 0. By re-arranging the code a little, I've made the
4065 decoding happen first, so such errors now appear on stderr, and the return
4066 code is 1. However, the actual setting up of the sockets still happens in
4067 the disconnected process, so errors there are still only recorded on the
4070 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4071 connections (as happens on some IP stacks) was logged at start up time as
4072 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4073 it from "IPv6 and IPv4", which means that two separate sockets are being
4076 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4077 says whether AF_INET or AF_INET6 was passed as an argument.
4079 42. Exiwhat output was messed up when time zones were included in log
4082 43. Exiwhat now gives more information about the daemon's listening ports,
4083 and whether -tls-on-connect was used.
4085 44. The "port" option of the smtp transport is now expanded.
4087 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4088 silently ignored. Now an error message is written to the main and panic
4091 46. There's a new ACL modifier called "logwrite" which writes to a log file
4092 as soon as it is encountered.
4094 47. Added $local_user_uid and $local_user_gid at routing time.
4096 48. Exim crashed when trying to verify a sender address that was being
4099 49. Exim was recognizing only a space character after ".include". It now also
4100 recognizes a tab character.
4102 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4103 extracting the relevant information from the specification. The man page no
4104 longer contains scrambled data for the -d option, and I've added a section
4105 at the front about calling Exim under different names.
4107 51. Added "extra_headers" argument to the "mail" command in filter files.
4109 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4112 53. Installed eximstats 1.29.
4114 54. Added transport_filter_timeout as a generic transport option.
4116 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4117 Cc: header lines. This was required by RFC 822, but it not required by RFC
4120 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4121 incoming messages that did not have them. Now it does so only if the
4122 message originates locally, that is, if there is no associated remote host
4123 address. When Resent- header lines are present, this applies to the Resent-
4124 lines rather than the non-Resent- lines.
4126 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4127 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4129 58. Messages for configuration errors now include the name of the main
4130 configuration file - useful now that there may be more than one file in a
4131 list (.included file names were always shown).
4133 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4134 for those rare installations that do not start the daemon as root or run it
4135 setuid root. I've cut out the call to initgroups() if the daemon is not
4138 60. The Exim user and group can now be bound into the binary as text strings
4139 that are looked up at the start of Exim's processing.
4141 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4143 62. Added $mailstore_basename variable.
4145 63. Installed patch to sieve.c from Michael Haardt.
4147 64. When Exim failed to open the panic log after failing to open the main log,
4148 the original message it was trying to log was written to stderr and debug
4149 output, but if they were not available (the usual case in production), it
4150 was lost. Now it is written to syslog before the two lines that record the
4151 failures to open the logs.
4153 65. Users' Exim filters run in subprocesses under the user's uid. It is
4154 possible for a "deliver" command or an alias in a "personal" command to
4155 provoke an address rewrite. If logging of address rewriting is configured,
4156 this fails because the process is not running as root or exim. There may be
4157 a better way of dealing with this, but for the moment (because 4.30 needs
4158 to be released), I have disabled address rewrite logging when running a
4159 filter in a non-root, non-exim process.
4165 1. The buildconfig auxiliary program wasn't quoting the value set for
4166 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4167 not defined. This bug was masked in 4.22 by the effect that was fixed in
4170 2. Some messages that were rejected after a message id was allocated were
4171 shown as "incomplete" by exigrep. It no longer does this for messages that
4172 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4174 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4175 have allow_domain_literals set, the ID did not get logged in the <= line.
4176 Domain literals are now always recognized in Message-ID: header lines.
4178 4. The first argument for a ${extract expansion item is the key name or field
4179 number. Leading and trailing spaces in this item were not being ignored,
4180 causing some misleading effects.
4182 5. When deliver_drop_privilege was set, single queue runner processes started
4183 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4184 same command in the process it spins off) were not dropping privilege.
4186 6. When the daemon running as "exim" started a queue runner, it always
4187 re-executed Exim in the spun-off process. This is a waste of effort when
4188 deliver_drop_privilege is set. The new process now just calls the
4189 queue-runner function directly.
4195 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4198 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4199 ignored. Though the use of -oP was forcing the writing of a pid file, it
4200 was always written to the default place.
4202 3. If the message "no IP address found for host xxxx" is generated during
4203 incoming verification, it is now followed by identification of the incoming
4204 connection (so you can more easily find what provoked it).
4206 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4208 5. Added some features to "harden" Exim a bit more against certain attacks:
4210 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4211 be put in Local/Makefile. This is like the never_users runtime option,
4212 but it cannot be overridden. The default setting is "root".
4214 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4215 prefix string with which any file named in a -C command line option
4218 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4219 is retained for -C and -D only if the caller of Exim is root. Without
4220 it, the exim user may also use -C and -D and retain privilege.
4222 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4223 command line option is disabled.
4225 6. Macro names set by the -D option must start with an upper case letter, just
4226 like macro names defined in the configuration file.
4228 7. Added "dereference=" facility to LDAP.
4230 8. Two instances of the typo "uknown" in the source files are fixed.
4232 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4233 the Configure-Makefile script screwed up while processing it.
4235 10. Incorporated PCRE 4.4.
4237 11. The SMTP synchronization check was not operating right at the start of an
4238 SMTP session. For example, it could not catch a HELO sent before the client
4239 waited for the greeting. There is now a check for outstanding input at the
4240 point when the greeting is written. Because of the duplex, asynchronous
4241 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4242 way, but not yet received, when the check is performed.
4244 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4245 on TCP/IP sockets, because this apparently causes some broken clients to
4248 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4249 unchanged) from the Cygwin maintainer.
4251 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4252 was supported instead of testing for mbx. Effectively a typo.
4254 15. The spa authenticator server code was not checking that the input it
4255 received was valid base64.
4257 16. The debug output line for the "set" modifier in ACLs was not showing the
4258 name of the variable that was being set.
4260 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4262 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4263 lookup on its own. Something else was needed to trigger the lookup. For
4264 example, a match in host_lookup or the need for a host name in a host list.
4265 Now, if $sender_host_name is referenced and the host name has not yet been
4266 looked up, a lookup is performed. If the lookup fails, the variable remains
4267 empty, and $host_lookup_failed is set to "1".
4269 19. Added "eqi" as a case-independent comparison operator.
4271 20. The saslauthd authentication condition could segfault if neither service
4272 nor realm was specified.
4274 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4275 error message that was logged was misleading, and incoming SMTP
4276 connections were dropped. The message is now more accurate, and temporary
4277 errors are given to SMTP connections.
4279 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4280 (except RSET) with a 421 error, until QUIT is received. However, it was
4281 failing to send a response to QUIT.
4283 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4284 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4285 ACL is now run after the helo_try_verify_hosts code.
4287 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4288 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4289 case-independent, so other case variants are also recognized). Apparently
4290 some systems use these upper case variants.
4292 25. If more than two messages were waiting for the same host, and a transport
4293 filter was specified for the transport, Exim sent two messages over the
4294 same TCP/IP connection, and then failed with "socket operation on non-
4295 socket" when it tried to send the third.
4297 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4299 27. The extern definition of crypt16() in expand.c was not being excluded when
4300 the OS had its own crypt16() function.
4302 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4303 as a preferred synonym for return_size_limit, both as an option and as an
4306 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4308 30. Changed the default configuration ACL to relax the local part checking rule
4309 for addresses that are not in any local domains. For these addresses,
4310 slashes and pipe symbols are allowed within local parts, but the sequence
4311 /../ is explicitly forbidden.
4313 31. SPA server authentication was not clearing the challenge buffer before
4316 32. log_message in a "warn" ACL statement was writing to the reject log as
4317 well as to the main log, which contradicts the documentation and doesn't
4318 seem right (because no rejection is happening). So I have stopped it.
4320 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4321 However, I am unable to do any testing of this.
4323 34. Fixed an infelicity in the appendfile transport. When checking directories
4324 for a mailbox, to see if any needed to be created, it was accidentally
4325 using path names with one or more superfluous leading slashes; tracing
4326 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4328 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4329 contents are added to the log line that is written for every discarded
4330 recipient. (Previously a log_message setting was ignored.)
4332 36. The ${quote: operator now quotes the string if it is empty.
4334 37. The install script runs exim in order to find its version number. If for
4335 some reason other than non-existence or emptiness, which it checks, it
4336 could not run './exim', it was installing it with an empty version number,
4337 i.e. as "exim-". This error state is now caught, and the installation is
4340 38. An argument was missing from the function that creates an error message
4341 when Exim fails to connect to the socket for saslauthd authentication.
4342 This could cause Exim to crash, or give a corrupted message.
4344 39. Added isip, isip4, and isip6 to ${if conditions.
4346 40. The ACL variables $acl_xx are now saved with the message, and can be
4347 accessed later in routers, transports, and filters.
4349 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4350 strings in the file are not string-expanded.
4352 42. If a MAIL command specified a SIZE value that was too large to fit into an
4353 int variable, the check against message_size_limit failed. Such values are
4354 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4355 day this will have to be increased, but I don't think I want to be around
4356 when emails are that large.
4363 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4364 iconv() is not standard in FreeBSD.
4366 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4367 IPv6 enabled. The observed symptom was a segmentation fault on return from
4368 the function os_common_find_running_interfaces() in src/os.c.
4370 3. In the check_special_case() function in daemon.c I had used "errno" as an
4371 argument name, which causes warnings on some systems. This was basically a
4372 typo, since it was named "eno" in the comments!
4374 4. The code that waits for the clock to tick (at a resolution of some fraction
4375 of a second) so as to ensure message-id uniqueness was always waiting for
4376 at least one whole tick, when it could have waited for less. [This is
4377 almost certainly not relevant at current processor speeds, where it is
4378 unlikely to ever wait at all. But we try to future-proof.]
4380 5. The function that sleeps for a time interval that includes fractions of a
4381 second contained a race. It did not block SIGALRM between setting the
4382 timer, and suspending (a couple of lines later). If the interval was short
4383 and the sigsuspend() was delayed until after it had expired, the suspension
4384 never ended. On busy systems this could lead to processes getting stuck for
4387 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4388 process, before it forks any delivery processes. The open lookup caching
4389 mechanism meant that the open file or database connection was passed into
4390 the delivery process. The problem was that delivery processes always tidy
4391 up cached lookup data. This could cause a problem for the next delivery
4392 process started by the queue runner, because the external queue runner
4393 process does not know about the closure. So the next delivery process
4394 still has data in the lookup cache. In the case of a file lookup, there was
4395 no problem because closing a file descriptor in a subprocess doesn't affect
4396 the parent. However, if the lookup was caching a connection to a database,
4397 the connection was closed, and the second delivery process was likely to
4398 see errors such as "PGSQL: query failed: server closed the connection
4399 unexpectedly". The problem has been fixed by closing all cached lookups
4400 in a queue runner before running a delivery process.
4402 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4403 seem to have the "const" qualifier which it has on other OS. I've
4406 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4407 *of the same type* (client or server) with the same public name that an
4408 error should be diagnosed.
4410 9. When Exim looked up a host name for an IP address, but failed to find the
4411 original IP address when looking up the host name (a safety check), it
4412 output the message "<ip address> does not match any IP for NULL", which was
4413 confusing, to say the least. The bug was that the host name should have
4414 appeared instead of "NULL".
4416 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4417 user that is built into the binary, and the -C or -D options is used, root
4418 privilege is dropped before the configuration file is read. In addition,
4419 logging is switched to stderr instead of the normal log files. If the
4420 configuration then re-defines the Exim user, the unprivileged environment
4421 is probably not what is expected, so Exim logs a panic warning message (but
4424 However, if deliver_drop_privilege is set, the unprivileged state may well
4425 be exactly what is intended, so the warning has been cut out in that case,
4426 and Exim is allowed to try to write to its normal log files.
4432 1. smtp_return_error_details was not giving details for temporary sender
4433 or receiver verification errors.
4435 2. Diagnose a configuration error if two authenticators have the same public
4438 3. Exim used not to create the message log file for a message until the first
4439 delivery attempt. This could be confusing when incoming messages were held
4440 for policy or load reasons. The message log file is now created at the time
4441 the message is received, and an initial "Received" line is written to it.
4443 4. The automatically generated man page for command line options had a minor
4444 bug that caused no ill effects; however, a more serious problem was that
4445 the procedure for building the man page automatically didn't always
4446 operate. Consequently, release 4.20 contains an out-of-date version. This
4447 shouldn't happen again.
4449 5. When building Exim with embedded Perl support, the script that builds the
4450 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4451 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4453 6. The freeze_tell option was not being used for messages that were frozen on
4454 arrival, either by an ACL or by local_scan().
4456 7. Added the smtp_incomplete_transaction log selector.
4458 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4459 was accepting AUTH without a new EHLO.
4461 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4462 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4465 10. Logging of TCP/IP connections (when configured) now happens in the main
4466 daemon process instead of the child process, so that the TCP/IP connection
4467 count is more accurate (but it can never be perfect).
4469 11. The use of "drop" in a nested ACL was not being handled correctly in the
4470 outer ACL. Now, if condition failure induced by the nested "drop" causes
4471 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4472 or "require"), the connection is dropped.
4474 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4475 that yield "discard" can now be used with an "accept" or a "discard" verb,
4476 but an error is generated for any others (because I can't see a useful way
4477 to define what should happen).
4479 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4480 are now processed in the same way as lines in the Exim configuration file.
4481 In particular, continuation lines are supported.
4483 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4485 15. Added -ti meaning -t -i.
4487 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4488 domains, and warn by logging if others are found.
4490 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4491 returned from GIFCONF when Exim is trying to find the list of interfaces on
4492 a host. The code in os.c has been modified to copy each ifreq to an aligned
4493 structure in all cases.
4495 Also, in some cases, the returned ifreq's were being copied to a 'struct
4496 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4497 means the last couple of bytes of an IPv6 address could be chopped if the
4498 ifreq contained only a normal sockaddr (14 bytes storage).
4500 18. Named domain lists were not supported in the hosts_treat_as_local option.
4501 An entry such as +xxxx was not recognized, and was treated as a literal
4504 19. Ensure that header lines added by a DATA ACL are included in the reject log
4505 if the ACL subsequently rejects the message.
4507 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4508 MD5 (which is deprecated).
4510 21. When testing a filter file using -bf, Exim was writing a message when it
4511 took the sender from a "From " line in the message, but it was not doing so
4512 when it took $return_path from a Return-Path: header line. It now does.
4514 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4515 with a valid header line field name (a series of printing characters
4516 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4518 23. Changed "disc" in the source to "disk" to conform to the documentation and
4519 the book and for uniformity.
4521 24. Ignore Sendmail's -Ooption=value command line item.
4523 25. When execve() failed while trying to run a command in a pipe transport,
4524 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4525 could be confused with a return value of 69 from the command itself. This
4526 has been changed to 127, the value the shell returns if it is asked to run
4527 a non-existent command. The wording for the related log line suggests a
4528 non-existent command as the problem.
4530 26. If received_header_text expands to an empty string, do not add a Received:
4531 header line to the message. (Well, it adds a token one on the spool, but
4532 marks it "old" so that it doesn't get used or transmitted.)
4534 27. Installed eximstats 1.28 (addition of -nt option).
4536 28. There was no check for failure on the call to getsockname() in the daemon
4537 code. This can fail if there is a shortage of resources on the system, with
4538 ENOMEM, for example. A temporary error is now given on failure.
4540 29. Contrary to the C standard, it seems that in some environments, the
4541 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4542 program. Exim now does this explicitly; it affects the formatting of
4543 timestamps using strftime().
4545 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4546 complaints. I've now initialized all the variables, to avoid this.
4548 32. Header lines added by a system filter were not being "seen" during
4549 transport-time rewrites.
4551 33. The info_callback() function passed to OpenSSL is set up with type void
4552 (*)(SSL *, int, int), as described somewhere. However, when calling the
4553 function (actually a macro) that sets it up, the type void(*)() is
4554 expected. I've put in a cast to prevent warnings from picky compilers.
4556 34. If a DNS black list lookup found a CNAME record, but there were no A
4557 records associated with the domain it pointed at, Exim crashed.
4559 35. If a DNS black list lookup returned more than one A record, Exim ignored
4560 all but the first. It now scans all returned addresses if a particular IP
4561 value is being sought. In this situation, the contents of the
4562 $dnslist_value variable are a list of all the addresses, separated by a
4565 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4566 to accept a host name and all its aliases if the forward lookup for any of
4567 them yielded the IP address of the incoming connection. Now it accepts only
4568 those names whose forward lookup yields the correct IP address. Any other
4569 names are discarded. This closes a loophole whereby a rogue DNS
4570 administrator could create reverse DNS records to break through a
4571 wildcarded host restriction in an ACL.
4573 37. If a user filter or a system filter that ran in a subprocess used any of
4574 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4575 the wrong values were passed to the pipe command ($thisaddress had the
4576 value of $0, $0 had the value of $1, etc). This bug was introduced by
4577 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4579 38. Improved the line breaking for long SMTP error messages from ACLs.
4580 Previously, if there was no break point between 40 and 75 characters, Exim
4581 left the rest of the message alone. Two changes have been made: (a) I've
4582 reduced the minimum length to 35 characters; (b) if it can't find a break
4583 point between 35 and 75 characters, it looks ahead and uses the first one
4584 that it finds. This may give the occasional overlong line, but at least the
4585 remaining text gets split now.
4587 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4588 file descriptors might be low, and that setting 1000 would always raise it.
4589 It turns out that in some environments, the limit is already over 1000 and
4590 that lowering it causes trouble. So now Exim takes care not to decrease it.
4592 40. When delivering a message, the value of $return_path is set to $sender_
4593 address at the start of routing (routers may change the value). By an
4594 oversight, this default was not being set up when an address was tested by
4595 -bt or -bv, which affected the outcome if any router or filter referred to
4598 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4599 log only when "message" or "log_message" are set. However, if one of the
4600 conditions was an address verification, or a call to a nested ACL, the
4601 messages generated by the underlying test were being passed through. This
4602 no longer happens. The underlying message is available in $acl_verify_
4603 message for both "message" and "log_message" expansions, so it can be
4604 passed through if needed.
4606 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4607 new expansion $bh_ to give the encoded byte string without charset
4608 translation. Translation happens only if iconv() is available; HAVE_ICONV
4609 indicates this at build time. HEADERS_CHARSET gives the charset to
4610 translate to; headers_charset can change it in the configuration, and
4611 "headers charset" can change it in an individual filter file.
4613 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4614 that creates RFC 2047 encoded "words" labels them as that charset instead
4615 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4616 expansion operator; (ii) when Exim creates a From: line for a local
4617 message; (iii) when a header line is rewritten to include a "phrase" part.
4619 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4620 buggy, causing it to skip the first lines of messages whose message ID
4621 ended in 'D'. This would not have bitten before Exim release 4.14, because
4622 message IDs were unlikely to end in 'D' before then. The effect was to have
4623 incorrect size information for certain domains.
4625 45. #include "config.h" was missing at the start of the crypt16.c module. This
4626 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4629 46. If there was a timeout during a "random" callout check, Exim treated it as
4630 a failure of the random address, and carried on sending RSET and the real
4631 address. If the delay was just some slowness somewhere, the response to the
4632 original RCPT would be taken as a response to RSET and so on, causing
4633 mayhem of various kinds.
4635 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4636 when I implemented it. It didn't allow for the fact that some option values
4637 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4638 the right test for positive values.
4640 48. Domain names in DNS records are case-independent. Exim always looks them up
4641 in lower case. Some resolvers return domain names in exactly the case they
4642 appear in the zone file, that is, they may contain uppercase letters. Not
4643 all resolvers do this - some return always lower case. Exim was treating a
4644 change of case by a resolver as a change of domain, similar to a widening
4645 of a domain abbreviation. This triggered its re-routing code and so it was
4646 trying to route what was effectively the same domain again. This normally
4647 caused routing to fail (because the router wouldn't handle the domain
4648 twice). Now Exim checks for this case specially, and just changes the
4649 casing of the domain that it ultimately uses when it transmits the message
4652 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4655 50. If a filter generated a file delivery with a non-absolute name (possible if
4656 no home directory exists for the router), the forbid_file option was not
4659 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4660 the existing equality matching.
4662 52. Exim was using ints instead of ino_t variables in some places where it was
4663 dealing with inode numbers.
4665 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4666 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4667 called TMPDIR, and if it finds it is different, it changes its value.
4669 54. The smtp_printf() function is now made available to local_scan() so
4670 additional output lines can be written before returning. There is also an
4671 smtp_fflush() function to enable the detection of a dropped connection.
4672 The variables smtp_input and smtp_batched_input are exported to
4675 55. Changed the default runtime configuration: the message "Unknown user"
4676 has been removed from the ACL, and instead placed on the localuser router,
4677 using the cannot_route_message feature. This means that any verification
4678 failures that generate their own messages won't get overridden. Similarly,
4679 the "Unrouteable address" message that was in the ACL for unverifiable
4680 relay addresses has also been removed.
4682 56. Added hosts_avoid_esmtp to the smtp transport.
4684 57. The exicyclog script was not checking for the esoteric option
4685 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4686 will work only if exicyclog is run under the appropriate euid.
4688 58. Following a discussion on the list, the rules by which Exim recognises line
4689 endings on incoming messages have been changed. The -dropcr and drop_cr
4690 options are now no-ops, retained only for backwards compatibility. The
4691 following line terminators are recognized: LF CRLF CR. However, special
4692 processing applies to CR:
4694 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4695 nor a local message in the state where . is a terminator.
4697 (ii) If a bare CR is encountered in a header line, an extra space is added
4698 after the line terminator so as not to end the header. The reasoning
4699 behind this is that bare CRs in header lines are most likely either
4700 to be mistakes, or people trying to play silly games.
4702 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4703 was being incorrectly given as 18 bytes larger than it should have been.
4704 This is a VOB (very old bug).
4706 60. This may never have affected anything current, but just in case it has:
4707 When the local host is found other than at the start of a list of hosts,
4708 the local host, those with the same MX, and any that follow, are discarded.
4709 When the list in question was part of a longer list of hosts, the following
4710 hosts (not currently being processed) were also being discarded. This no
4711 longer happens. I'm not sure if this situation could ever has previously
4714 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4717 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4718 header. This is something that is recommended in a new Internet Draft, and
4719 is something that is documented as being done by Sendmail. There are two
4720 possible values. For messages generated by the autoreply transport, Exim
4723 Auto-Submitted: auto-replied
4725 whereas for all other generated messages (e.g. bounces) it adds
4727 Auto-Submitted: auto-generated
4729 63. The "personal" condition in filters now includes a test for the
4730 Auto-Submitted: header. If it contains the string "auto-" the message it
4731 not considered personal.
4733 64. Added rcpt_include_affixes as a generic transport option.
4735 65. Added queue_only_override (default true).
4737 66. Added the syslog_duplication option.
4739 67. If what should have been the first header line of a message consisted of
4740 a space followed by a colon, Exim was mis-interpreting it as a header line.
4741 It isn't of course - it is syntactically invalid and should therefore be
4742 treated as the start of the message body. The misbehaviour could have
4743 caused a number of strange effects, including loss of data in subsequent
4744 header lines, and spool format errors.
4746 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4747 client host had authenticated. This control can now be exercised by an ACL
4748 for more flexibility.
4750 69. By default, callouts do not happen when testing with -bh. There is now a
4751 variant, -bhc, which does actually run the callout code, including
4752 consulting and updating the callout cache.
4754 70. Added support for saslauthd authentication, courtesy of Alexander
4757 71. If statvfs() failed on the spool or log directories while checking their
4758 size for availability, Exim confusingly gave the error "space shortage".
4759 Furthermore, in debugging mode it crashed with a floating point exception.
4760 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4761 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4762 serious problem, Exim now writes to the main and panic logs when this
4763 happens, with details of the failure. It then refuses to accept the
4764 incoming message, giving the message "spool directory problem" or "log
4765 directory problem" with a 421 code for SMTP messages.
4767 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4768 0, 1, and 2 exist, because some OS complain for execs without them (see
4769 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4770 descriptors. However, the code omitted to check that the open succeeded,
4771 causing mysterious errors if for some reason the permissions on /dev/null
4772 got screwed. Now Exim writes a message to the main and panic logs, and
4773 bombs out if it can't open /dev/null.
4775 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4776 interact so that it is all more flexible. It is supposed to remain
4777 backwards compatible. Also added extra_local_interfaces.
4779 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4780 to bomb out with an assertion failure - to the client this appears as a
4781 connection drop. This problem occurs in the part of the code that was taken
4782 from the Samba project. Fortunately, the assertion is in a very simple
4783 function, so I have fixed this by reproducing the function inline in the
4784 one place where it is called, and arranging for authentication to fail
4785 instead of killing the process with assert().
4787 75. The SPA client code was not working when the server requested OEM rather
4788 than Unicode encoding.
4790 76. Added code to make require_files with a specific uid setting more usable in
4791 the case where statting the file as root fails - usually a non-root-mounted
4792 NFS file system. When this happens and the failure is EACCES, Exim now
4793 forks a subprocess and does the per-uid checking as the relevant uid.
4795 77. Added process_log_path.
4797 78. If log_file_path was not explicitly set, a setting of check_log_space or
4798 check_log_inodes was ignored.
4800 79. If a space check for the spool or log partitions fails, the incident is now
4801 logged. Of course, in the latter case the data may get lost...
4803 80. Added the %p formatting code to string_format() so that it can be used to
4804 print addresses in debug_print(). Adjusted all the address printing in the
4805 debugging in store.c to use %p rather than %d.
4807 81. There was a concern that a line of code in smtp_in.c could overflow a
4808 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4809 initially expressed, the concern was not well-founded, because trailing
4810 spaces are removed early. However, if the trailing spaces were followed by
4811 a NULL, they did not get removed, so the overflow was possible. Two fixes
4814 (a) I re-wrote the offending code in a cleaner fashion.
4815 (b) If an incoming SMTP command contains a NULL character, it is rejected
4818 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4819 runs initgroups(), so that if the Exim user is in any additional groups,
4820 they will be used during message reception.
4826 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob